No title

1. Introduction

Innovative pricing effects: theory and practice in mobile Internet networks Andreas Jonason

The author Andreas Jonason is Senior Price Analyst at Sitarg Consulting and Post-doctoral Researcher at the Royal Institute of Technology, Stockholm, Sweden. Keywords Financial innovations, Competences, Pricing strategy, Mobile communications, Internet, Strategy Abstract Multidimensional products, such as telecommunications, often distinguish themselves by a considerable number of spillovers. The creator of the spillovers wants to commercialize the spillovers but cannot if the parameters on which the ability to charge is proprietary to another firm in the economy. These actors then need to agree upon an efficient pricing contract to be able to charge. As an effect, the direction of revenue in the transaction may not be self-evident, turning previous customers into suppliers and vice versa. The paper presents data from the mobile Internet market to validate this claim empirically and make suggestions of possible solutions to the pricing problem facing these actors with the introduction of mobile multimedia. Electronic access The research register for this journal is available at http://www.emeraldinsight.com/researchregisters The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/1460-1060.htm

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . pp. 185±193 # MCB UP Limited . ISSN 1460-1060 DOI 10.1108/14601060210451144

A central element within microeconomic and management theory, notably principal agent theory (e.g. Hart and Holmstrom, 1987; Holmstrom and Milgrom, 1991) and value chain theory (e.g. Porter, 1985) is the concept of the exogenous relationship between actors. The principal always knows he is a principal, the agent always knows he is an agent. Pricing in these models is restricted to maximize revenue minus cost (microeconomics), the maximization of value for a given customer (e.g. Porter, 1980, 1985) or the utilization of monopoly elements for market advantages (e.g. Goodman and Lawless, 1994). Regardless of its value, however, a product will not be a success (in terms of profit) unless its producer (owner) can make it chargeable in the market. The problem of making products chargeable becomes acute if decision models become subject to uncertainty (as defined by Knight, 1921) and continuously change as a result of market dynamics. This situation arises when products are multidimensional and the producer has difficulties determining which dimensions are in demand and which dimensions can be charged for. In such a situation the product or output (Hayek, 1945) is not well defined and, hence, neither is price. Given such a situation, firms often tend to search within their product, or adopt their products endogenously, to find a stable element to make the whole chargeable, as opposed to making it more valuable (through product innovation) through a search process called innovative pricing (IP) ( Jonason, 1999, 2001). It will be shown how the effect of this search can reconfigure existing pricing institutions and actually redefine the organizations (Hardaker et al., 2000). Hence the so often taken-for-granted roles in the market come up for question. How should then actors, dealing in multidimensional products, go about reaching an efficient pricing The author would particularly like to thank Dr Sven Modell, Prof. Uwe Cantner, Prof. Stan Metcalfe, Prof. Clas Wihlborg and Prof. Gunnar Eliasson for their assistance in the completion of this paper. In addition, the author wants to thank Ê ke Persson, Tapio Antilla, Jonas Colleen Wade, A Lindh, Anders Olin and Anders Henriksson for comments, support and suggestions that have been very valuable. Finally the author wants to thank Prof. Pervaiz K. Ahmed and two anonymous reviewers for their comments and suggestions on earlier versions of this article.

185

Innovative pricing effects: theory and practice in mobile Internet networks

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 185±193

Andreas Jonason

wireless infrastructure and the three operators in question. Section 7 summarizes the key features of the paper, and makes some recommendations for corporate pricing decisions. For the purpose of this paper the operators will remain anonymous.

contract? This question introduces our problem.

2. The problem Consider the following simple example: private builders and owners of infrastructure, notably bridges, are often faced with a pricing problem: according to theory users should be charged their marginal external cost (Pigou, 1920; Knight, 1924). One problem, however, is that the valuable element of the product (the bridge) does not contain a chargeable element (pricing base). Instead, the solution to the pricing problem is to endogenously expand the product (the bridge) to include a pricing base (for instance a toll booth), that does not make the product more valuable from an end-user perspective, but makes it chargeable. The problem treated in this paper concerns the situation when the owner of the valuable entity (the bridge)[1] and the owner of the chargeable entity (the toll booth) are two separate firms and, therefore, need to agree upon a pricing contract. Technically more complex products are faced with this type of pricing problem such as the delivery of content and applications to mobile devices, typically news, banking and entertainment services[2]. The providers of the content and applications contribute value to the service, but rely on the network operators to charge the end-user. It is shown how the operators tend to be myopic in their pricing contracts towards the providers and thus reduce their incentives to continue to contribute value. The spillovers in the market, coupled with the pricing base, make the roles in the market endogenous and subject to dynamics. The problem is illustrated theoretically, both in terms of evolutionary economics and management control theory, creating a bridge between the two. The remainder of this paper is structured as follows. To begin with, the theory of innovative pricing (IP) and its place in strategic literature is presented. Section 4 argues how the institutionalized pricing scheme between agents can influence the definition of the roles of buyer and seller. The analysis then draws on empirical data to demonstrate these points. Three different mobile Internet revenue models are reviewed, each from a different operator. The data were gathered in cooperation with Ericsson Radio Systems, the world's largest supplier of

3. An innovative pricing strategy Strategists, such as Porter (1980, 1985), have developed a highly regarded paradigm of strategy centralized around the creation of value. Another common focus of strategy (e.g. Goodman and Lawless, 1994) suggests that firms should seek a position in which they can exhibit a measure of monopolistic power. The creation of value and the benefits of monopolistic power are obviously of central strategic importance to any business but, as argued here, need to be accompanied by the issue of chargeability. Market power or amounts of value created are irrelevant unless the firm can find a way to transform created value into revenue. Hence, 1, value, 2. monopolistic control (property rights) and 3, chargeability are interdependent. By experimenting with different pricing combinations in the market, a firm might strike gold and find a pricing solution that allows it to charge not only for the value it creates but in addition take advantage of gratis value created by other actors in the economy (i.e. the spillovers). The firm may in fact decide to modify its pricing base of a product in an attempt to capture spillovers created by other actors in the economy, rather than by its own products. Through this the firm uses its innovative competence in pricing to endogenously adapt the product definition to a pricing situation rather than, the more common academic problem, the opposite. Conventional price theory often deals with homogenous goods or bundles thereof (e.g. Tirole, 1988; Wilson, 1993; Ramsey, 1927; Mitchell and Vogelsang, 1991). Product definitions and pricing structures are often considered exogenous, however, and not as part of a path dependent, subjective and innovative decision process of the firm. In such models the externalities between products are known or at least calculable under some rational expectations function. The problem, however, is that when the number of combinations available to the firm are numerous, and change quickly over time, it prevents a pricing decision

186

Innovative pricing effects: theory and practice in mobile Internet networks

Andreas Jonason

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 185±193

based on rational expectations, since the optimization upon which the mainstream pricing theory rests cannot incorporate an open state space[3]. Schumpeterian economics with its focus on the role of the entrepreneur is therefore more helpful. The entrepreneur is a person who develops new products not seen previously in the market. Thus Schumpeter (1911) views innovation as the identification of commercially viable inventions. The change in science is only interesting in so far as it is capable of transforming the outside world, and that this change has to be commercially successful. An innovation, hence, is not only the creation of something new but also involves the ability to act commercially on the opportunities that this advance in science and technology offers. This returns us to the issue of pricing as an essential tool of innovation and commercial success. It may well be the case that the pricing scenario is given while the product definition, in terms of function, cost or quality is not. The pricing problem will often go hand in hand with the product definition. This integrates pricing and product innovation and will, hence, open up an evolutionary price/product setting called IP ( Jonason, 1999). Management control literature sometimes brings up the related concept of the institutional entrepreneur as a breaker of institutional frameworks, such as pricing. First introduced by DiMaggio (1988) and later applied to institutional theory by, for instance, Fligstein (1997) and Kondra and Hinings (1998) the institutional entrepreneur is portrayed as an important contributor to cost reductions within the firm. Such studies often address the duality of institutional rules and practices pointing to the inability of static rules to give a unanimous answer to how agents should act. Gupta and Govindarajan (1984) state that firms, facing a high level of uncertainty, experience higher performance by placing greater reliance on subjective evaluation rather than on calculated mechanical (i.e. optimized) and institutionalized responses. There are, hence, gains to be made in constantly questioning, subjectively expanding and reorganizing the charging of products. Such IP business experiments thus continuously test existing institutional practices. An institutional entrepreneur in a strategic typology is comparable to the theory developed by Miles and Snow (1978) and Miller and Friesen (1982). It distinguishes between

conservative defenders and entrepreneurial prospectors. The defining variable is the rate at which the firm alters its products and markets relative to its competitors. Defenders are organizations that have a narrow and stable product line. Such agents tend to adopt institutionally stable routines. Like Schumpeter's (1911) and DiMaggio's (1988) institutional entrepreneur, the prospector, in conflict with the defender, constantly examines the environment for product and market opportunities. Prospectors rely on the product and market development to bring change to the environment around them to generate profits, and therefore face a more uncertain environment (Miles and Snow, 1978). What are these subjective pricing experiments based on? Beckert (1999) addresses this issue by defining institutions as reducers of uncertainty by creating expectations in the behavior of others. The precondition to deducing strategies is for each agent to know how other relevant actors in the field will act. In accordance with Schumpeter (1911), Beckert distinguishes the role of the manager as a follower of existing institutional rules from that of the entrepreneur as a breaker of such takenfor-granted rules while searching for strategic opportunities to enhance self-interest. In a similar manner, one solution to the problem of how to act in the face of uncertainty is to create institutional practices around certain subjectively stable product elements. In regards to the IP problem this entails identifying a proprietary chargeable characteristic of the product that subjectively relates to the rest of the product portfolio in a predictable way. Such elements could, for instance, be parameters that are of a monopolistic nature. This stable characteristic can then be leveraged to charge for additional elements (spillovers) within the product portfolio that create value but less revenue than their marginal cost. Such biased parameters have to exist, or be discovered (invented), however, before the product can be made chargeable. The effect could include a redefinition (invention of a new) of the pricing institution in line with Beckert's argument. The firm can, through the identification and redefinition of stable parameters, leverage its pricing to charge for parameters to suit its purposes, and thereby create expectations among other actors of how, and at which price level, similar products will be priced in the future. The

187

Innovative pricing effects: theory and practice in mobile Internet networks

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 185±193

Andreas Jonason

innovative pricer thrives in such an environment creating and redefining the pricing institutions in the market contributing to reduced uncertainty ``creating the mechanism that allows the economy to work'' (Day, 1984). The entrepreneurial prospector is an innovative pricer, a breaker and a creator of institutional pricing practices by being dynamic in leveraging stable characteristics of its changing product portfolio, transforming the creation of value and the monopolistic elements into revenues through endogenous expansion and redefinition of its pricing base.

4. Ripples by IP As argued above, the theory of IP entails finding a stable characteristic to make the valuable elements of the product chargeable. An innovation does not have to be spectacular (Schumpeter, 1911). An IP scheme can be something that is brought in from another market as long as it entails something new that other actors through inability or ignorance cannot exploit. For instance in the previous example a solution to the bridge-pricing problem could simply be to set up a toll booth[4]. An interesting problem arises, however, when the owner of the stable chargeable characteristic and the owner of the valuable entity are two separate actors in the economy. In other words when the owner of the bridge and the toll booth are two separate firms. An instant questions arises: who should pay whom? The answer, even for physical goods, may not be that self-evident given the presence of externalities. Take mobile phones, for instance, that are given away for free, charged at, and sometimes even charged below, marginal cost depending on the chargeable externalities they create in network traffic (airtime), branding, etc. This type of problem has been addressed for non-price innovations. For instance, in their popular microeconomic work, Nelson and Winter (1982) bring attention to the point that an innovation is often produced by a firm for sale to customers who will use it. There are two acts of innovation involved in this context (buyer and seller). With pricing being an innovation both agents in the transaction need to have a properly working feedback mechanism on how different pricing schemes

affect overall demand to make the spillovers created chargeable. DiMaggio's (1988) institutional entrepreneur will need to be placed on both sides of the transaction. A vital ingredient in the commercial success of the IP experiment is, therefore, the receiver competence of the customer, understanding and contributing not only to the development and improvement of the product (Eliasson and Eliasson, 1996) but also to its pricing. Suppliers are often restricted in their ability to affect the pricing of their customers and the effects can, hence, be out of their control. This can create an incentive to redefine their product, to reduce uncertainty, either to exit the provider market or entry in the operator market through vertical integration or by partnering with substitutes. Both outcomes are obviously non-beneficial, and this naturally raises the interesting question of how to avoid this scenario. The customer's role is fulfilled partly as an informant of specifications that are needed for the products to remain competitive, for instance on how the product should be defined in terms of price. This can be arranged to create a positive sum game, where both parties can come out as entrepreneurs and winners. Failure to work together will blunt the innovation or at least subject one of the parties to uncertainty that can be used by the competition to gain an advantage. Both parties taking part in a transaction can thus be innovative pricers. This fits well with Williamson's (1988) fundamental transformation, which states how the bidding condition is transformed into bilateral supply. Because of this transformation opportunism can create high transaction costs if dealt in the usual market mode. It may then be in the interest of all parties, both owning the specialized asset (the ability to charge, the operator) and those renting or buying the asset (the provider) to join each other in vertical integration (Klein et al., 1978). The marketing strategy to be efficient, hence, may have to include active support to the customer on how to structure the pricing scheme to capture spillovers, or to include an element of joint ownership of the pricing base (revenue sharing). If suppliers benefit from a particular pricing scheme they may be willing to stimulate and even subsidize a particular product definition to their customers, or even to their customers' customers deliberately losing money on it if it stimulates the demand of a chargeable, and perhaps unique, externality. The important

188

Innovative pricing effects: theory and practice in mobile Internet networks

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 185±193

Andreas Jonason

element of such an IP scheme is a feedback mechanism that the customer understands and can charge for these externalities (intangible cloud of spilled value), such that the supplier is properly rewarded, by revenues, for the value it creates. An important source of information in this discovery will be those firms benefited (potentially and actively) by the spillovers acting as informants and giving rewards for the spillovers created (Eliasson, 2001). This redefinition of customer/supplier relationships is continuous and depends on the general pricing dynamics and on how firms decide to improve their receiver competence. The customer/supplier definition is based on a subjective interpretation of the externalities in the market. The firm will need to discover which customers are actually customers and which should be paid (turned into suppliers) an above marginal cost price for the externalities they create. The theoretical hypothesis to bring with us into the empirical section of this paper is thus first to show the endogenous roles in the market as a result of how the actors decide to charge. Second, to use the different pricing models as benchmarks of efficiency to give suggestions of how the pricing contract and the feedback mechanism could be set up in order to avoid the pricing problem some of the operators are facing.

5. Mobile Internet definition through pricing One industry that has experienced an abrupt change in its pricing institution is wireless telecommunications. From being a regular voice service device with its pricing based on a per minute of use base (airtime), mobile phones now offer a wide range of contributions from several industries with multiple voice, content and application plans. This has created uncertainty not only about how these services are charged for but also how the pricing of the existing services (primarily voice), still dominant in revenue, relates to the new. Many operators have initiated mobile Internet portals that offer comparably simple push and e-mail services to their subscribers. Since content services and applications are new to most operators,

many have decided to outsource this role to third party providers that already hold a fixed Internet presence. The most common arrangement is that the operators handle the billing towards the end-users, while the providers add value to the end-users in the form of branded services and applications. It is debated, however, how end-users' revenues are to be divided between these third party providers and the operators. The providers would like to see themselves as suppliers of services, and therefore expect to receive payment. The operators do not always agree, however. Providers are often completely dependent on the operators' billing system to make their products chargeable, a monopolistic element that the operators often take advantage of. A positive externality created from the services is a form of branding for the providers (or general market growth[5]) which could increase the demand for products sold, and made chargeable, elsewhere (the demand for Disney films, for instance). The operators, on the other hand, receive the positive externality of airtime for transporting the services in their networks. To return to the previous example of the bridge and the toll booth. The providers add the value (bridge) but depend on the operators (toll booth) to make this value chargeable. As illustrated below, however, the decision on how to define customers and suppliers (direction of revenue in the transaction) will differ between markets depending on the operators' subjective perception of the spillovers created. The role of the providers as either customer (paying a positive price) or supplier (paying a negative price) is endogenous and as illustrated below in Figure 1 varies between markets. The first indications from the mobile Internet market point to a considerable advantage for the wireless operators being able to leverage their monopoly on end-user access and charging to gain the lion share of revenues (Jonason and Eliasson, 2001). At the same time, providers are clearly an important element in the creation of end-user value, airtime usage, and a potential diversifier for the operators against the competition. Technically there are still considerable improvements on the horizon (such as mobile broadband). Given the institutional effects of

189

Innovative pricing effects: theory and practice in mobile Internet networks

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 185±193

Andreas Jonason

IP, however, it is likely that the pricing institutions of today will set a standard of how mobile Internet pricing contracts are to be arranged in the future.

6. Observations The empirical part of this study was carried out at Ericsson Radio Systems, the world's largest supplier of wireless infrastructure, between June 1999 and December 2000. Three types of actors in the mobile Internet industry were studied: (1) service providers; (2) end-users; and (3) operators. The service providers included providers of content and applications. End-users were mobile Internet subscribers, i.e. traditional customers of the operators. The structure of the analysis has been focused around the role of the third type of agents, operators. Three operators were selected, with a special focus on two of these (Nos. 1 and 2). Open-ended interviews with executives at the operators and Ericsson internally were initially conducted to gain an overall understanding of how the pricing schemes were set up in relation to the different actors. Another important source of information was actual billing data published by the operators themselves. These data were then verified against the findings of Ericsson's own consumer lab research, consisting of several rounds of structured interviews with over 2000 end-users from relevant domestic markets. The findings were presented and discussed with nine different operators, several service providers and a considerable number of internal Ericsson groups for verification and feedback.

7. Results The operators' pricing strategies were classified in Figure 1. It is important for the reader to note, however, that these pricing models are simplified for the purpose of illustrating the points of this paper. An example of this is the exclusion of terminal subsidies, which can have a considerable effect on airtime usage. An additional exception is that both operators 1 and 2 have a combined flat-fee with their

incremental end-user charge. Finally, operators 2 and 3 also have exceptions to their general rule of charging their providers (some actually do get paid). The models are also dynamic in the sense that the pricing contracts between the actors constantly changed during the course of the study. These simplifications do not change the general snapshot picture illustrated by Figure 1, however. The three operators in the figure all have different ways of charging end-users. Operator 1 charges for the information (data traffic) sent and received by counting the number of data packets. Operator 2 uses a similar strategy but, because of technical limitations, charges its end-users on a circuit-switched basis, in minute increments. These two operators have a considerable dependency on airtime as a base for charging. If competition forces lower prices on airtime (by the introduction of 3G?), it may force these operators to exit the market or redefine their bases for charging to make money elsewhere. Finally, operator 3 uses a flat fee for charging endusers. This pricing scheme has advantages in end-user simplicity and a cost-efficient billing process. However, there is also considerable risk associated with network utilization in terms of managing capacity. Under this pricing scheme, operator 3 may be forced to upgrade capacity to cover the demand of existing customers and thus drive up network costs without a corresponding revenue increase. Under the regime imposed by operators 2 and 3, the providers are in fact customers (paying a positive price for placement on the portal menu). Many of the same providers are at the same time paid by operator 1 (or more correctly the end-users of operator 1 through revenue sharing) to deliver similar content and services. This makes the operator the customer. The problem in models 2 and 3 is that there is little revenue incentive for the providers. The mobile Internet service of these operators may actually work as a substitute to the providers' fixed Internet presence. The pricing decision of the operators has, hence, helped to structure the roles of the different agents in the market (supplier or customer). A more long-term solution for these providers can be to either stop offering the services (exit the market) or to become operators themselves, by for instance buying airtime or

190

Innovative pricing effects: theory and practice in mobile Internet networks

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 185±193

Andreas Jonason

Figure 1 Three different mobile Internet pricing models (simplified)

even bidding for a 3G license[6] (redefinition through vertical integration). The latter option would surely turn the providers into competitors rather than partners, where the competitive advantage could be reached by leveraging their unique content or applications. The three pricing scenarios presented, although snapshots, illustrate the constant market experiments in the mobile Internet market. Operators and providers constantly change their product and pricing definitions in their evolutionary search for profits. A popularized pricing scheme can quickly become a standard. For this purpose, pricing is conceived as a routine and potentially institutionalized transaction practice. Institutionalized means that pricing, over time, can come to underpin the ``taken-for-granted'' ways of thinking in a particular product/market setting. Pricing solutions can remain static until an entrepreneur (innovative pricer) finds a way to turn the market upside down, again, creating a new pricing institution; since the technical changes and the unpredictability of the decisions of other actors in the economy, feeds a dynamic process of continuously changing bases for pricing. The pricing institution as defined by the innovative pricer will, hence, need to be adaptive (Metcalfe, 1988) to changes in the environment. Just as with any technical standard the continuous problem for a firm is when to

abandon a particular pricing scheme. Firms are always at risk of using an outdated method of charging[7]. The decision of how to charge does not only affect the competition, the suppliers and the customers of the product. It is also a major defining element of the actors' roles in the transaction (customer or supplier). As illustrated, the providers can be both customers and suppliers to the operators. Operators 2 and 3 have defined the providers as customers (i.e. paying a positive price) while operator 1 defines its providers as suppliers (i.e. paying a negative price). The set standard pricing scheme is maintained by the transaction costs and institutions of the economy and hinders deviation from what is considered normal, until an entrepreneurial prospector comes up with a new IP scheme that turns the whole market upside down again. Another conclusion is that previously partners either deliberately or unwittingly can become competitors if the charging scheme is covered in secrecy or if it blocks them from profits in the market. The operators' decision of how to price their mobile Internet presence will therefore not only be a tool of competition but also a major determinant of which companies they will be able to regard as partners and which will become competitors. As illustrated, the problem with this type of externality pricing towards providers is that it reduces the incremental incentives to provide a compelling and at the same time low cost

191

Innovative pricing effects: theory and practice in mobile Internet networks

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 185±193

Andreas Jonason

(little airtime usage) content service for the end-users of the service. Unless providers are committed to offering a compelling service it will reduce the externality of airtime that makes the service chargeable for the operator, the receiver competence may consequently have to be built. Inability of suppliers and providers to charge for products can have two effects. They can either exit the market or redefine their products, which potentially could make them competitors. None of these are beneficial for operators, since this either would lessen the overall value of the end product or subject the operators to increased competition. The ability for suppliers and partners to efficiently charge for their products, hence, becomes a joint problem for all parties. The critical path of the mobile Internet business is, hence, to create a mutual understanding of the value of the product such that the right incentive contract can be drawn. The providers in the mobile Internet model will have to depend on revenues from elsewhere. These revenues can come from building a stronger fixed Internet brand through its mobile presence or by bypassing the operators' access to the end-customers by becoming an operator themselves, changing their product and pricing base to enable chargeability (IP). Therefore, a perhaps more beneficial pricing scheme for the operators may be to share bases for pricing with their providers, on airtime for instance, to align the revenue incentives of all participants. The pricing scheme of the firm has to allow partners to be profitable and informed of upcoming pricing schemes if the firm wants them to remain as partners. This means that firms should share pricing information or pricing base (revenue sharing), often regarded secret and sacred, to reduce uncertainty in the market (i.e. improve the receiver competence of customers) and align the incentives of all parties. In a Beckert (1999) setting this would enable a beneficial institutional framework for charging the mobile Internet services of the future.

decisions affect not only the institutional rules of the firm but also the rules (i.e. the range of possible pricing strategies) available to other agents in the value chain. They are endogenous in the market. Not only partners can be turned into competitors. Suppliers can also be redefined to become customers. The conclusion emphasizes the rationale of sharing information and incentives on pricing strategies with both suppliers and customers. By sharing information on its pricing strategy the firm increases the receiver competence of its customers and thus reduces uncertainty of how to price. Furthermore, a shared pricing base with suppliers, (i.e. revenue sharing) is likely to reduce the incentive for the supply side to bypass the role of the firm through IP (inventing an alternative base for charging) and may thereby lower the business risk for both parties.

Notes

8. Conclusions This paper has drawn on theoretical and empirical material to bring elements of management control, strategy and microeconomic pricing together. Pricing 192

1 Those readers who question the multidimensionality of the bridge example can have a look at the impact the Oresund bridge has had on real estate prices in the Malmoe-Copenhagen region. Should the builders of the bridge be rewarded for the property value they created? 2 Examples of providers are Yahoo!, Amazon, Bandai and Disney. 3 A segment of the CDMA division of Ericsson Radio Systems, for instance, which is only a part of the infrastructure side of Ericsson, has over 14,000 different products in its portfolio. The number of bundling combinations for pricing is, hence, 314,000 (exceeding the expected number of particles in the universe). The rapid technical development in telecommunications contributes to making the demand externalities between these bundles dynamic over time (the turnover rate is approximately two years). A mainstream Neoclassicist would, hence, need to explore a state space considerably larger than the universe that changes in definition over a period of two years, before he could even start his maximization process of revenue minus cost. Needless to say, this is very far from what goes on when Ericsson Radio Systems is pricing its products. 4 This solution may not be very spectacular. In Schumpeterian theory it is the novelty of the innovation that counts. The first person to think of a toll booth to charge for the usage of a bridge was hence clearly an entrepreneur (innovative pricer). 5 For instance, one of Japan's major gas companies offers free recipes as a mobile content service. The motivation for the zero price is that free recipes stimulate the usage of gas for heating food (rather than eating raw fish). 6 Or in the case of Sweden, appearing beautiful enough to receive a free UMTS license from the government.

Innovative pricing effects: theory and practice in mobile Internet networks

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 185±193

Andreas Jonason

7 See, for instance, Jonason (1999) for examples from the telecommunications market, where the old PTT Telia was rightfully described as outdated in its pricing of fixed line telephony.

References Beckert, J. (1999), '`Agency entrepreneurs, and institutional change: the role of strategic choice and institutional practices in organizations'', Organization Studies, Vol. 20 No. 5, pp. 777-99. Day, R. (1984), ``Disequilibrium economic dynamics: a post Schumpeterian approach'', Journal of Economic Behavior and Organization, Vol. 5 No. 1, pp. 57-76. DiMaggio, P.J. (1988), ``Interest and agency in institutional theory'', in Zucker, L. (Ed.), Institutional Patterns and Organizations, pp. 3-32, Ballinger, Cambridge, MA. Eliasson, G. (2001), ``Innovative pricing and surreptitious spillovers'', paper presented to the 28th Annual EARIE Conference, Dublin, September. Eliasson, G. and Eliasson, AÊ. (1996), ``The biotechnical competence block'', Revue d'Economie, Industrielle, Vol. 4 No. 78, Trimestre. Fligstein, N. (1997), ``Social skill and institutional theory'', American Behavioral Scientist, Vol. 40, pp. 397-405. Goodman, R.A. and Lawless, M.W. (1994), Technology and Strategy Conceptual Models and Diagnostics, Oxford University Press, New York, NY. Gupta, A.K. and Govindarajan, V. (1984), ``Business unit strategy, managerial characteristics, and business unit effectiveness at strategy implementation'', Academy of Management Journal, pp. 25-41. Hardaker, G., Ahmed P.K. and Graham, G. (2000), ``Non-linear innovation across the Internet'', Integrated Manufacturing Systems, Vol. 7 No. 11, pp. 484-90. Hart, O.D. and Holmstrom, B. (1987), ``The theory of contracts'', in Bewley, T. (Ed.), Advances in Economic Theory, Cambridge University Press, New York, NY. Hayek, F. (1945), ``The use of knowledge in society'', The American Economic Review, Vol. 15 No. 4, pp. 21-30. Holmstrom, B. and Milgrom, P. (1991), ``Multitask principal agent analyses: incentive contracts, asset ownership and job design'', Journal of Law, Economics, and Organizations, No. 7, pp. 24-52. Jonason, A. (1999), '`Trading in, and pricing of not easily defined products and services'', in Eliasson, G. (Ed.), Internet Economics and Internet Technology.

Jonason, A. (2001), ``Innovative pricing, universitetsservice'', PhD thesis, Royal Institute of Technology. Jonason, A. and Eliasson, G. (2001), ``Mobile Internet revenues: an empirical study of the I-mode portal'', Journal of Internet Research, No. 4. Klein, B., Crawford, R.G. and Alchian, A.A. (1978), ``Vertical integration, appropriable rents, and the competitive contracting process'', Journal of Law and Economics, No. 21, pp. 297-326. Knight, F. (1921), Risk, Uncertainty and Profits, Houghton-Mifflin, Boston, MA. Knight, F. (1924), ``Some fallacies in the interpretation of social cost'', Quarterly Journal of Economics, Vol. 38, pp. 582-606. Kondra, A.Z. and Hinings, C.R. (1998), ``Organizational diversity and change of institutional theory'', Organization Studies, Vol. 5 No. 19, pp. 743-67. Metcalfe, J.S. (1988), ``Evolution and economic change'', in Silberston, A. (Ed.), Technology and Economic Progress, Macmillan, Basingstoke. Miles, R.E. and Snow, C.C. (1978), Organizational Strategy, Structure and Process, McGraw-Hill, New York, NY. Miller, D. and Friesen, P.H. (1982), ``Innovation in conservative and entrepreneurial firms: two models of strategic momentum'', Strategic Management Journal, pp. 1-25. Mitchell, B.M. and Vogelsang, I. (1991), Telecommunications Pricing: Theory and Practice, Cambridge University Press, Cambridge. Nelson, R. and Winter, S. (1982), An Evolutionary Theory of Economic Change, Harvard University Press, Cambridge, MA. Pigou, A.C. (1920), Wealth and Welfare, Macmillan, London. Porter, M.E. (1980), Competitive Strategy, Free Press, New York, NY. Porter, M.E. (1985), Competitive Advantage, Free Press, New York, NY. Ramsey, F.P. (1927), ``A contribution to the theory of taxation'', Economic Journal, No. 37, pp. 47-61. Schumpeter, J. (1911), The Theory of Economic Development, (English ed., 1934), Harvard Economic Studies, Vol. XLVI, Harvard University Press, Cambridge, MA. Tirole, J.T. (1988), The Theory of Industrial Organization, MIT Press, Cambridge MA. Williamson, O.E. (1988), ``Economics and sociology: promoting a dialog'', in Farkas, G. and England, P. (Eds), Industries, Firms and Jobs: Sociological and Economic Approaches, Plenum, New York, NY, pp. 159-85. Wilson, R. (1993), Nonlinear Pricing, Oxford University Press, London.

193

New product development in German and US technology firms James T. Simpson Christine Kollmannsberger Helmut Schmalen and David Berkowitz The authors James T. Simpson is Professor of Marketing and Director at the Center for the Management of Science and Technology, The University of Alabama in Huntsville, Huntsville, Alabama, USA. Christine Kollmannsberger is Co-owner and Vice President at Kollmannsberger KG, Ingolstadt, Germany. Helmut Schmalen is University Professor at the University of Passau, Passau, Germany. David Berkowitz is Assistant Professor of Marketing at the University of Alabama in Huntsville, Huntsville, Alabama, USA. Keywords Germany, USA, New product development, High technology, National cultures, Marketing Abstract This research tested a model in both Germany and the USA that contained marketing variables known to impact new product development success in high technology firms. We explore the link between national culture and new product development. A multi-group LISREL analysis revealed that while the model structure is valid for both countries, the impact of certain marketing factors on commercial product success differed. The analysis revealed that the mean values of the marketing factors differed significantly in each country. The most important determinant of new product success in both countries is proficiency in product launch activities. Marketing skills also positively impact product success in both countries. Top management support and involvement was also an important predictor of new product success in Germany, but the impact was negative. The potential impact of culture on the findings is discussed. Electronic access The research register for this journal is available at http://www.emeraldinsight.com/researchregisters The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/1460-1060.htm European Journal of Innovation Management Volume 5 . Number 4 . 2002 . pp. 194±207 # MCB UP Limited . ISSN 1460-1060 DOI 10.1108/14601060210451153

New product development success is critical to firm survival regardless of the firm's geographic location. Moreover, culture plays an important role in developing the operational norms and the rules of behavior in the firm. Therefore, it is interesting to note that marketing researchers have made few attempts to directly link national culture and new product development (Nakata and Sivakumar, 1996). Numerous researchers have both identified and investigated the antecedents of new product development success (c.f. Lynn et al., 1997; Song and Parry, 1997; Montoya-Weiss and Calantone, 1994). Yet, Nakata and Sivakumar (1996) were the first to develop an integrative review and conceptual model of the relationship between culture and new product development. Previous research on culture has focused on many diverse issues with respect to the relationship between culture and its impact on the management of the firm (Clark, 1990). It has been found that culture can often dictate firm structure and operational norms (Nakata and Sivakumar, 1996). Many studies designed to determine the relationship between culture and marketing outcomes have been conducted over the last several decades (c.f. Green and Langeard, 1975; Tse et al., 1988). Nevertheless, few attempts have been made to investigate the impact of culture on the organization's new product development plans and strategies. The failure to integrate culture into new product development studies is largely responsible for the absence of a culturally generic model of new product development. The objective of this paper is to report on a national study of new product development in high technology firms in the USA and Germany. Specifically, we address the link between successful technology and market oriented new product development. We develop, from the literature, a non-culturally specific conceptual model of the new product development process, and then we test the model in two distinct cultures. Clark (1990) points out that cross-cultural studies currently have no integrating theory. Thus, this exploratory study proposes to fuel the theory building process by providing descriptive The authors would like to thank Professor William E. Souder for his support of this project.

194

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

results that should enhance our understanding of how new product development processes differ in countries with dissimilar cultures.

Conceptual background Culture There are several definitions of culture. Linton (1945, p. 32) defined culture as a ``configuration of learned behaviors and results of behavior whose component parts are shared and transmitted by members of a particular society''. Hofstede (1991 p. 4) offered the following definition of culture, ``the collective programming of the mind that distinguishes the members of one category of people from those of another''. In a very important study, Hofstede (1980) identified four dimensions of culture. This seminal work was based on a survey of managers from one firm across 53 national cultures. Hofstede's (1980) four dimensions included: (1) power distance; (2) uncertainty avoidance; (3) individualism; and (4) masculinity. Hofstede provides numerical scores on each of these dimensions for each culture. From a separate study of 23 countries, Bond et al. (1987) proposed the Confucian dynamic as a fifth cultural dimension. Dimensions of culture Power distance Power distance is the extent to which the less powerful members of institutions accept and expect that power is distributed unequally (Hofstede and Bond, 1988). Societies scoring high on power distance tend to be more autocratic and are more accepting of variation in wealth and power than societies with lower scores. On the power distance dimension, both Germany and the USA have similar and relatively low scores. Both countries place a high value on equality and democratic participation. The USA is only slightly more tolerant of unequally distributed power (Hofstede, 1980). Uncertainty avoidance Uncertainty avoidance generally relates to how well societies deal with the ambiguities of the future. Specifically, this dimension

captures the extent to which societies feel threatened by ambiguity and create institutions that attempt to minimize uncertainty (Hofstede and Bond, 1988). On the uncertainty avoidance dimension, the USA and Germany are relatively similar with Germany scoring at the mid-range of all cultures and the USA scoring in the lower range. The implication is that US cultured employees are somewhat more tolerant of uncertainty than German employees (Hofstede, 1980). Individualism vs collectivism The notion of individualism and collectivism refers to the relationship between the individuals in a society and the collective that exist in a society (Hofstede, 1980). These two dimensions can be presented as the ends of a bipolar continuum. At one end, individualism is a situation in which members of a society are loosely aligned and look after only themselves and their immediate family. At the opposite end, collectivism is a situation in which members adopt group norms and seek collective interests (Hofstede and Bond, 1988). On this dimension, there is a large disparity between the two countries. While Germany falls in the mid-range of all countries, the USA scored higher than all other countries in the study. The USA is a very individualist society while Germany tends to be more of a collectivist society (Hofstede, 1980). Masculinity vs femininity Masculinity exists when the society is characterized as assertive and the dominant values are success, money, and material things. On the other hand, femininity exists when the society is characterized as nurturing and the dominant values are caring for others and quality of life (Hofstede and Bond, 1988). On this dimension, the scores for the USA and Germany are almost identical. Both countries tend to be more masculine. Confucian dynamism The term Confucian dynamism evolved from a study conducted by Bond and his colleagues (1987) that used a Chinese value survey (CVS) to assess cultural differences across 23 countries. The CVS study identified a fifth dimension, which parallels the values of Confucianism. On the dynamic side are future oriented values such as thrift, hard

195

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

work and perseverance. On the opposite side are present and past oriented values such as respect for tradition and reciprocation. Since the Confucian dynamic is also relevant to countries without a Confucian past, Hofstede (1991) preferred to refer to the Confucian dynamic as a long-term versus short-term orientation. On this dimension, both Germany and the USA scored in the midrange of all countries. The scores were very similar with Germany assigning slightly more importance to future oriented values (Hofstede and Bond, 1988). Based on an analysis of these five important cultural dimensions, there are only minimal differences between Germany and the USA. Both countries are very similar on power distance, masculinity and Confucian dynamism. The USA is somewhat more tolerant of uncertainty than Germany. The major difference in the two societies is that the USA is more loosely aligned and concerned primarily with the needs of the individual while Germany is more concerned about the needs of the group. In addition, Glunk et al. (1996) reported a series of culturally embedded management practices that differed between German and American managers. Model development and hypotheses The objectives of this study require the development of a parsimonious model that can be tested in two separate cultures. Thus, we attempt to identify a literaturebased model that contains marketing variables known to impact new product development success. The development of our model was influenced by previous studies in new product development (NPD) (c.f. Song and Parry, 1997; Brown and Eisenhardt, 1995; Montoya-Weiss and Calantone, 1994). Our conceptual model (Figure 1) is most similar to one proposed by Song et al. (1997). Their model proposed that five factors: (1) process skills; (2) project management skills; (3) alignment of skills with needs; (4) team skills; and (5) design sensitivity lead to marketing and technical proficiency, which in turn determine product quality

and, ultimately, new product success or failure. Our model argues that commercial product success (CPS) is moderated by proficiency in the product development initiation (PPD) stage, which is influenced by technical congruence (TC), project manager style (PMS), and market congruence (MC). In addition, commercial product success (CPS) is moderated by proficiency in the product launch (PPL) stage, which is influenced by project manager style (PMS) and marketing skills (MS). There is also a direct influence from proficiency in the initiation (PPD) stage to proficiency in product launch (PPL). Finally, we suggest that both top management involvement and support (TMS) for the project and the project team's level of motivation (MOT) directly impact new product success. Definitions of these variables can be found in the measurement section. The following hypotheses can be derived from the proposed model: H1. A project team's proficiency in the initiation stage of the product development (PPD) process is related positively to the firm's perception of new product commercial success (CPS). H2. A project team's proficiency in the product launch stage of the new product development process (PPL) is related positively to the firm's perception of new product commercial success (CPS). H3. The project team's proficiency in the initiation stage of the product development process (PDP) is related positively to proficiency in the product launch stage (PPL). H4. The compatibility between the project team's technical skills and technical requirements of the project (TC) are positively related to proficiency in the initiation stage of the product development process (PDP). H5. The project manager's participatory style (PMS) is related positively to proficiency in the initiation stage of the product development process (PDP).

196

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

Figure 1 Factors leading to new product commercial success

H6. The project manager's participatory style (PMS) is related positively to proficiency in the product launch stage (PPL). H7. The compatibility between the firm's existing markets and the new product requirements (MC) is positively related to the proficiency in product development (PPD). H8. The firm's marketing skills (MS) are related positively to proficiency in product launch (PPL). H9. Top management support (TMS) for the project is related positively to the firm's perception of new product commercial success (CPS). H10. The team's motivation (MOT) is positively related to the firm's perception of commercial product success (CPS). It is expected that all relationships presented in the preceding hypotheses will be positive and significant for both the German and US data. Yet, it is likely that the magnitude of the relationships in the model might differ by country. We expect that differences in magnitude among relationships can be attributed to cultural differences between the two countries.

Method Product samples and data collection Data were collected from US and German high technology firms ( Joint Economic Committee of Congress, 1981) producing pharmaceutical, computer, electronics, software and transportation products. German and US firms were matched based on size, R&D intensity, and financial performance. A range of firm sizes was selected in both countries. Products for the study were selected by the chief executive officers at each firm, in consultation with the authors. The products were required to be new to each firm (not simply a new version of an older product), involve significant development and marketing activities for that firm, and have detailed life cycle information available. Products were also required to have sufficient market experience so that their degree of commercial product success (CPS) could be adequately judged. The sampling design specified two success products and two failed products with equal representation from all industries (i.e. pharmaceutical, software). Ultimately, 151 products were selected. A total of 21 successful products and 19 failed products were selected from the German

197

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

firms, while 58 successful products and 53 failed products were selected from the US firms. The smaller German sample reflects the relative populations of German firms available for this study, the loss of firms resulting from matching US and German firms, and the availability of products meeting the study requirements. All the questionnaires were completed through personal interviews that obtained 100 percent response rates to all items and allowed the collection of supporting anecdotal information.

proficiency of product launch measure assesses the project team's perception of its proficiency in developing manufacturing, marketing, and technical service start-ups.

Instrument development and pilot testing The initial questionnaire for this study was formulated in the English language and pilot tested in the USA. Two bilingual German nationals using a double-back translation process developed the German version. The questionnaires used in this study were designed to collect data on the ``representative values'' of the antecedents during the life cycle of the NPD efforts. To achieve this, respondents were instructed to reflect on the most representative condition (value) of each antecedent over the life of the NPD effort and rate it accordingly. Measures Dependent variables The primary dependent variable is degree of commercial product success or failure. This single item measure is designed to assess the extent to which the commercial outcome of the project is consistent with the firm's prior expectations with expectations defined in terms of original goals set for the product. The goals were specified in terms of sales, market share, return on investment, profit, customer satisfaction, contribution to technology leadership and contribution to market leadership, in both the USA and Germany. The two additional endogenous variables are proficiency in the initiation stage and the launch stage of the product development process. Proficiency in product development is a four-item scale designed to measure the project team's perception of how proficiently it performs the exploratory, concept development, prototype development, and testing stages of the new product development process. The three-item

Independent variables The five-item technical congruence measure captures the extent to which the firm's R&D, engineering and manufacturing skills are both at the appropriate level and compatible with the project requirements. The four items' project manager style measure captures the project team member's perception of how well the project manager employs a participative decision-making process and attempts to motivate team members. Five items describing the project team's competencies in market research, sales, and demand forecasting measure marketing skills. Three items capture the project team perception of top management's involvement and support for the project. Similar to technical congruence, the three-item marketing congruence scale measures the similarities among the firm's existing markets, product lines and marketing skills, and the new project requirements. Finally, the project team's level of motivation was assessed with a single-item global measure. A sample of items used for each measure is presented in Table I. Measure validation for the US-German sample The two-step measure validation process involved pooling the US and German data into one 151-product database (Anderson and Gerbing, 1982). The first step includes a three-phase scale validation for the measurement model. First, each multi-item scale was subjected to item-to-total correlation analysis. Correlations below 0.5 were excluded from further analysis. Next, principal component factor analysis was used to assess the dimensional qualities of each scale. Two items that impaired internal consistency or cross-loaded were eliminated. Of the 28 remaining items, 26 items are associated with the seven multi-item measures and two items are associated with two single-item measures. The final step of measure validation involves subjecting the multi-item measures to confirmatory factor analysis using LISREL 8 ( Joreskog and Sorbom, 1993).

198

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

Table I Response formats and sample items Measure(s)

Response anchora

Sample item

Commercial success (CPS)

Far below expectations Far above expectations Very poor . . . very good

Proficiency in product development (PPD) Proficiency in product launch (PPL) Project manager style (PMS)

Very poor . . . very good

Technical congruence (TC) Marketing congruence (MC)

Strongly disagree . . . strongly agree Strongly disagree . . . strongly agree

Marketing skills (MS) Top management support (TMS) Motivation (MOT)

Strongly disagree . . . strongly agree Strongly disagree . . . strongly agree

How would you describe the commercial outcomes (i.e. profit, ROI, sales, market share) of this product? How proficiently were the prototype testing activities executed? How proficiently were the manufacturing start-up activities performed? There was adequate participation in the decision making by the people working on the project Our R&D skill was at the desired level for this project There was a close fit between our existing markets and the market for this product Our forecast of market demand for this product was accurate Top management was supportive of this project

Very low . . . very high

The motivation of the project level personnel was . . .

Strongly disagree . . . strongly agree

a

Note: All were five-point scales

Due to the relatively small ratio of observations per variable, four different measurement models composed of similar constructs were chosen for confirmatory analysis. The two single indicator measures were excluded from this analysis. The LISREL results presented in Table II reveal an acceptable level of measure reliability and validity for the pooled American and German data. The second step in the Anderson and Gerbing approach is to sum the indicators for each latent variable to form a new composite measure. To account for measurement error, the error variances were set to 0.05 for the two single-item measures. For the seven multi-item measures, the error variances were set to 1 minus the scale composite reliability. Structural model and analysis procedures The first objective of this research is to assess the effects of the success factors on new product success in the USA and Germany. The structural model presented in Figure 1 reflects the hypotheses detailed in Table III. The second objective is to assess the moderating effects of country (i.e. USA or Germany) on the nomological relationships among the variables in the model. It is anticipated that the magnitude of the relationships in the model might differ by country. More specifically, it is anticipated that the structural coefficients in our model might differ in size by country, but not in direction. Baron and Kenny (1986)

recommend multi-group analysis within LISREL to model the moderator variable effects in a structural model when independent data samples are available for each group.

Results Comparing the structural model for USA and Germany A test of equality of the two structural models (see Sauer and Dick, 1993) revealed that the US and German models have the same form or structure but different parameter values. There is substantial evidence that the US and German models have the same number of influencing variables with the same relationships among the variables. Still, the test suggests that the parameter values may differ in sign, size, and significance levels. Differences in parameter estimates and significant levels for each relationship in each country are presented in Table IV. The results presented in Table IV and Figure 2 support H1 for the German data but not for the US data. Project team proficiency in product development is positively related to product success in Germany. There is no significant relationship between proficiency in product development and success for US teams. There is strong support for H2 in both the USA and Germany. Project team proficiency in product launch is positively related to product success in both countries.

199

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

Table II Measurement items and measure validation Indicator reliability

Measure

Measurement model 1 Top management support (TMS) X1 X2 X3

0.64 0.51 0.60

Project manager style (PMS) X4 X5 X6 X7

0.28 0.52 0.48 0.75

Composite reliability

Variance extracted

0.81

0.60

0.80

0.51

0.84

0.57

0.76

0.52

0.78

0.50

0.71

0.46

0.90

0.64

Notes: c2/df = 0.88; GFI = 0.980; AGFI = 0.953; RMR = 0.46

Measurement model 2 Marketing skills (MS) X8 X9 X10 X11

0.70 0.76 0.42 0.40

Marketing congruence (MC) X12 X13 X14

0.44 0.30 0.83

Notes: c2/df = 0.25; GFI = 0.997; AGFI = 0.987; RMR = 0.11

Measurement model 3 Proficiency in product development (PPD) X15 X16 X17 X18

0.10 0.45 0.80 0.70

Proficiency in product launch (PPL) X19 X20 X21

0.27 0.33 0.58

Notes: c2/df = 1.43; GFI = 0.972; AGFI = 0.928; RMR = 0.37

Measurement model 4 Technical congruence (TC) X22 X23 X24 X25 X26

0.68 0.85 0.48 0.52 0.67

Motivationa (MOT) X27 Successa (CPS) X28 Notes: c2/df = 0.42; GFI = 0.999; AGFI = 0.983; RMR = 0.06; a single item measures

200

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

Table III Hypotheses and results by country Country Hypotheses

H1. A project team's proficiency in the initiation stage of the product development process (PPD) is related positively to the firm's perception of new product commercial success (CPS) H2. A project team's proficiency in product launch stage of the new project development process (PPL) is related positively to the firm's perception of new product commercial success (CPS) H3. The project team's proficiency in the initiation stage of the product development process (PDP) is related positively to proficiency in the product launch stage (PPL) H4. The compatibility between the project team's technical skills and technical requirements of the project (TC) are positively related to proficiency in the initiation stage of the product development process (PDP) H5. The project manager's participatory style (PMS) is related positively to proficiency in the initiation stage of the product development process (PDP) H6. The project manager's participatory style (PMS) is related positively to proficiency in the product launch stage (PPL) H7. The compatibility between the firm's existing markets and the new product requirements (MC) is positively related to the proficiency in product development (PPD) H8. The firm's marketing skills (MS) are related positively to proficiency in product launch (PPL) H9. The management support (TMS) for the project is related positively to the firm's perception of new product commercial success (CPS) H10. The team's motivation (MOT) is positively related to the firm's perception of commercial product success (CPS) a

USA

Germany

Not significant

Supported

Supported

Supported

Supporteda

Not supportedb

Not significant

Supported

Supported

Supported

Not significant

Not significant

Supported

Not significant

Supported

Supported

Not significant

Not supportedb

Supported

Not significant

b

Note: Significant at < 0.10, all others significant at < 0.05; sign is negative

Table IV Common metric completely standardized solution USA Parameter B21 B31 B32 Y11 Y12 Y13 Y22 Y24 Y35 Y36

PPD ! PPL PPD ! CPS PPL ! CPS TC ! PPD PMS ! PPD MC ! PPD PMS ! PPL MS ! PPL TM ! CPS MOT ! CPS

Estimates 0.39 ±0.16 0.51 ±0.12 0.64 0.47 0.05 0.46 ±0.02 0.53

t-value **

1.83 ±1.02 3.32* ±1.01 6.43* 4.01* 0.27 3.35* ±0.27 5.16*

German Estimates ±0.88 0.52 0.84 0.48 0.57 0.26 0.26 1.11 ±0.37 ±0.24

t-value ±3.06* 3.11* 4.78* 3.14* 4.41* 1.75 1.04 5.08* 2.83* ±1.41

Notes: * significant at " < 0.05; ** significant at " < 0.10; fit statistics for all models in both groups; chi-square/df quotient: 74.92/43 = 1.74; GFI = 0.800; CFI = 0.940; RMR = 0.13; RMSEA = 0.071

As proposed in H3, US project team proficiency in product development is positively related to proficiency in product launch. Counter to our expectations, this relationship is negative and significant in Germany. Taken collectively, these results

indicate a basic difference in the US and German antecedents to commercial product success. There is mixed support for H4. In Germany, the congruence between the project team's technical skills and the 201

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

Figure 2 Results of hypotheses tests

technical requirements of the project is positively related to proficiency in product development. However, this relationship is not significant in the USA. Consistent with H5, project manager style is positively related to proficiency in product development in both Germany and the USA. The results do not support H6. The project manager's participatory style is not related significantly to the team's proficiency in product launch in either country. There is partial support for H7. The congruence between the US project team's marketing capabilities and the requirements of the project is positively related to the US team's proficiency in product development. The same relationship for German teams is not significant. The results strongly support H8. The team's marketing skills are related positively to proficiency in product launch in both the USA and Germany. The relationship between top management support and success (H9) is not significant for US firms. Surprisingly, top management support in German firms is negatively related to commercial success. Finally, the project team's motivation is related positively to product success (H10) in US firms. Team motivation has no impact on success in German firms.

Decomposing the total influences into direct and indirect effects The preceding results reveal the direct effects of one variable on another as proposed by the model. Path analysis also allows us to decompose the direct and indirect effects of each antecedent variable on the dependent variable. The direct effects are the influences of one variable on another that are not mediated by any other variable (e.g. MOT ! CPS). Indirect effects apply when there is at least one intervening variable between the two variables of interest (e.g. MS ! PPL ! CPS). Decomposing the total effects (TE) into their direct (DE) and indirect (IDE) components identifies the most important success factors separately for both countries and can reveal additional patterns that explain differential findings across the two countries. Table V presents the direct, indirect and total effects based on non-standardized paths for each of the success factors. In the USA, three factors were identified with significant total effects on commercial product success (CPS). Proficiency of product launch (PPL) has the greatest total impact (TE = 0.60), the project team's motivation (MOT) is next (TE = 0.53), and marketing skills (MS) has the lowest impact (TE = 0.24). Since there are no intervening variables between either proficiency of

202

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

Table V Determinants of new product commercial success Non-standardized path estimates Direct effects Indirect effects Total effects

Antecedent variables

US firms Proficiency in product development (PPD) Proficiency in product launch (PPL) Technical congruence (TC) Project manager style (PMS) Marketing skills (MS) Top management support (TMS) Market congruence (MC) Motivation (MOT)

±0.17** 0.60* 0.00 0.00 0.00 ±0.03 0.00 0.53*

German firms Proficiency in product development (PPD) Proficiency in product launch (PPL) Technical congruence (TC) Project manager style (PMS) Marketing skills (MS) Top management support (TMS) Market congruence (MC) Motivation (MOT)

0.54* 0.98* 0.00 0.00 0.00 ±0.39* 0.00 ±0.24

Notes: *Significant at " < 0.05;

**

0.21** 0.00 0.00 0.05 0.24* 0.00 0.02 0.00 ±0.78* 0.00 ±0.11 0.10 0.95* 0.00 ±0.06 0.00

0.04 0.60* 0.00 0.05 0.24* ±0.03 0.02 0.53* ±0.24 0.98* ±0.11 0.10 0.95* ±0.39* ±0.06 ±0.24

Significant at " < 0.10

product launch (PPL) and success (CPS) or project team motivation (MOT) and success (CPS), the direct effects equal the total effects in these two cases. Thus, increases in proficiency of product launch (PPL) and project team motivation (MOT) substantially increase product success. However, marketing skills (MS) positively impact (IDE = 0.24) commercial success (CPS) through product launch (PPL). Other variables in the model have no total influence on commercial success (CPS) in the USA. There are also three factors that significantly impact on success in German firms. Like the USA, proficiency in product launch (PPL) has the most significant direct effect on product success (CPS) (TE = 0.98) and marketing skills (MS) positively impact success through product launch (PPL) (TE = IDE = 0.95). Unlike the USA, and contrary to our expectations, top management support (TMS) negatively impacts success (CPS) in German firms (DE = TE = ± 0.39). The decomposition of effects shown in Table V reveals substantially different patterns of relationships among some variables in the USA and Germany. For example, the total effects of proficiency in product development (PPD) on success (CPS) are non-significant (p > 0.10) in both

the USA and Germany. Yet, the decomposition of the total effects in both countries reveals other patterns. Product development proficiency (PPD) has a direct negative effect (DE = ±0.17, p < 0.10) on product success (CPS) in the USA. Yet, the indirect effect of product development proficiency (PPD), through product launch proficiency (PPL), is positively related to success (IDE = 0.21). These two significant but contradictory effects lead to a non-significant total effect (TE = 0.04, p > 0.10). A comparison of the data for these variables in Tables IV and V reconciles this seemingly contradictory result, revealing that the indirect effects consist of two component relationships. The first relationship is between product development proficiency (PPD) and product launch proficiency (PPL). This relationship (Table IV) is positive and marginally significant (0.39, p < 0.10). The second component is the relationship between product launch proficiency (PPL) and success (CPS). This relationship (Table IV) is positive and highly significant. Thus, the strong positive impact of product launch (PPL) on success (CPS) overpowers the indirect effect of product development (PPD) on success (CPS). Thus, when the impact of product launch proficiency (PPL) is eliminated, an increase in product

203

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

development proficiency (PPD) leads to less success (CPS). When one considers the strong mediating effect of proficiency in product launch (PPL), proficiency in product development (PPD) leads to greater success (CPS). A similar phenomenon exists in Germany, except that the signs are reversed and the relationship among variables is different. In Germany, the relationship between product development (PPD) and product launch (PPL) is dominant in determining the negative indirect effect of product development proficiency (PPD) on success (CPS). In other words, product development proficiency (PPD) has a significant negative influence on success (CPS) because of its dominant negative effect on product launch (PPL). The variance explained in this model (80 percent for Germany and 60 percent for the USA) further supports the notion that the model depicts a reasonable representation of factors that determine success in both the USA and Germany. Differences in mean values of success factors and intercepts The final analysis determined whether the latent success factor means differed significantly across countries. The means were estimated and then the analysis was repeated with the restriction of invariance across countries. A statistically significant (p < 0.05) chi-square increase of 12.6 with six degrees of freedom was obtained, suggesting that all success factors are not equal in Germany and the USA. Table VI reports the differences in mean values for each success factor. Positive values reflect a larger mean in Germany; negative values reflect a larger mean in the USA. All differences are significant at p < 0.05. Table VI Differences between the US and German means of the success factors Success factor Technical congruence (TC) Project manager style (PMS) Marketing skills (MS) Top management support (TMS) Market congruence (MC) Motivation (MOT)

An analysis of both the differences in mean values and previously reported parameter estimates facilitates a more comprehensive analysis. First, both technical and market congruence exists to a greater extent in the US sample. This finding suggests, that compared to German firms, US firms may concentrate more on developing products that are compatible with the firm's existing markets and technologies. Nevertheless, neither significantly impacts success in either the USA or Germany. Second, project manager style is slightly higher in Germany, but project manager style has no substantial impact on success in either country. Third, marketing skills are significant determinants of success in both countries. Yet, marketing skills are rated higher in Germany than in the USA. The higher level of marketing skills in Germany may explain why marketing skills is the second most important determinant of success in German firms and only the third most important in US firms. Fourth, the level of top management support and involvement is lower in Germany than in the USA. Yet, the involvement of top management does not significantly impact success in the USA. In Germany, top management support is the second most important predictor of success, but the impact is negative. This might suggest that in the USA the norm is for top managers to be substantially involved in all new product developments. In German firms, however, the norm may imply less top management involvement except in instances where products are likely to fail. Further aggravating the situation is the notion that German subordinates tend to almost always defer to the recommendations of top management. Finally, project team motivation is higher in German firms, but it is only a significant success determinant in US firms.

Difference to US meana

Discussion and implications

±0.30* 0.06* 0.34* ±0.30* ±0.23* 0.30*

Notes: *Significant at < 0.05; a Negative values reflect larger in the USA, positive values reflect larger in Germany

This study addresses the fundamental question of whether culture impacts the relationships between selected marketing factors and new product success in high technology firms. The study tested a model in both Germany and the USA that contained marketing variables known to impact new

204

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

product development success. A multi-group LISREL analysis revealed that while the model structure is valid for both countries, the impact of certain marketing factors on commercial product success differed. Further, the analysis revealed that the mean values of the marketing factors differed significantly in each country. The authors suggest that cultural backgrounds and management practices help to explain these differences: No matter how hard man tries, it is impossible for him to divest himself of his own culture . . . people cannot act or interact at all in any meaningful way except through culture (Hall, 1966 p.177).

Therefore, for teams to operate in an environment that has accepted cultural norms, they need to be aware of the impact of these norms on their performance. As a consequence, these cultural norms have different levels of absolute and relative impact. The most important information about effective cultural influence can be obtained from a total effect analysis. This analysis identifies the factors most responsible for the commercial success of new products in each country. In both countries only three of the latent variables have a significant total effect. Proficiency in product launch (PPL) is the most important predictor of new product success in both the USA and Germany. Further, the firms' marketing skills (MS) significantly impact the level of proficiency in product launch in both countries. Moreover, both the positive impact of marketing skills on PPL and the positive impact of PPL on commercial success are more profound in Germany than in the USA. A detailed analysis shows that marketing skills exist at a higher level in Germany than the USA. These results reveal the first cultural difference between the two countries. The marketing skill factor consists of variables such as competence in market research, market forecasting, and predicting customer requirements. A firm's commitment to the development of these skills reflects the firm's attitude towards ambiguities and uncertainty. Germany's higher score on uncertainty avoidance suggests that Germany is more anxious about the future than the USA. Germany, therefore, potentially places more emphasis on market research and forecasting skills as a way to

minimize market uncertainty. This finding is consistent with Ingelfinger (1995). The increased attention to market research by German firms leads to both reduced feeling of uncertainty and enhanced marketing skills. These enhanced marketing skills are associated with greater proficiency in product launch and consequently greater product success. The third variable with a significant total effect in Germany is top management support (TMS) and in the USA it is project team motivation (MOT). In Germany we observed a negative relationship between top management support and new product success. One possible explanation for this surprising result is rooted in both German management's technical orientation and German society's more collective nature. German top management may have very specific expectations about how he/she wants the product to perform technically. Yet, he/she may have little insight into the external environment or the implementation requirements of new product development. There is a tendency in Germany to not question those in authority, so poor suggestions will likely be implemented without question (Ingelfinger, 1995; Keller, 1987). The results suggest that the more German top management become involved in project team activities, the less likely the product will be commercially successful. For the more individualistic American project teams, top management support has no significant impact on commercial product success. Our results also suggest that the absolute level of top management support is significantly higher in the USA than in Germany. Collectively, these findings suggest that the relationship between top management support and product success might be non-linear, precisely u-shaped. Positive relationships may be found only at the extreme high and low values. Negative relationships between top management support and product success might be found at the mid-range values, as is the case in German firms. The level of top management support in the more individualistic US firms has surpassed the mid-range, but has not reached a level high enough to positively impact product success. Increased project team motivation (MOT) leads to increased product success in US

205

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

firms; yet, there is no relationship between team motivation and product success in Germany. Increased motivation is known to lead to increased effort and performance. Thus, the different relationships between motivation and success in the USA and Germany is likely unrelated to culture, but rather it is related to the different levels of motivation in each country. The absolute level of motivation in Germany is significantly higher than in the USA. It appears that the level of motivation in Germany is sufficiently high that marginal changes in motivation do not impact success given the other variables in the model. Beyond the total effect analysis, there are additional relationships described in the model that indicate the impact of culture on new product development. One of the more interesting findings relates to the impact of proficiency in the initiation stage of new product development on proficiency in the launch stage. As proposed, greater proficiency in the early stage of product development is associated with greater proficiency in the launch stage in US firms. Surprisingly, increases in proficiency in the initiation stage are associated with lower proficiency in the launch stage in German firms. One possible explanation is rooted in the supposition that there is a temporal dimension to most product introductions. Project teams frequently commit to deliver a new product to customers or participate in a trade show by a certain date. Further, project teams have finite resources. The technically oriented German managers tend to be highly detail oriented and committed to building as much quality into a product as possible. Thus, they may become so preoccupied with the early stage of the development cycle that they leave too little time for launch activities. In addition, the project team's enthusiasm for the technical quality of the product may lead the team to assume that the product's superiority diminishes the need for aggressive launch activities. In other words, German project teams likely spend a disproportionate amount of time on initiation to the detriment of the launch activities. This argument is similar to the one proposed by Ingelfinger (1995). The project management style (PMS) factor includes variables such as extent of team member participation in decisions and delegation of authority, which are related to

the power distance cultural dimension. Our findings show that the influence of this factor on PPD is the same for both countries and the mean value in Germany is only slightly higher than in the USA. These results are not surprising since Germany and the USA have similar and relatively low scores on the power distance dimension. Our findings suggest that culture might impact the extent to which the USA and Germany value compatibility between existing and future projects. Germany's lower mean scores on both technical and market congruence suggest that they are less concerned than the USA with how well their current technical and market competencies match those required by the proposed market. Germany's slightly higher score on the Confucian dynamic suggests that German firms are more oriented towards the future and would likely concentrate less on congruencies. US firms tend to be slightly more short-term oriented. Thus, higher levels of technical and market congruence would reduce time to market. Consistent with the notion that German firms are more technically oriented and US firms are more market oriented, increased levels of technical congruence lead to increased proficiency in product development only in German firms. Whereas, increased levels of market congruence lead to more proficient product development only in US firms. With the globalization of business, many firms are coordinating new product development efforts in different countries. Managerial implication This paper demonstrates that culture and management practices impact the new product development process. First, the importance of certain variables in determining new product success differs somewhat in Germany and the USA. The most important determinant of new product success in both countries is proficiency in product launch activities. Increasing project team motivation appears to be more important in the USA than in Germany. In fact, German project team members are on average more highly motivated than US team members, and increases in motivation have little impact on product success. Since product launch activities and marketing skills have a substantial impact on new product success in

206

New product development in German and US technology firms

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 194±207

J.T. Simpson, C. Kollmannsberger, H. Schmalen and D. Berkowitz

Germany, managers should allocate limited resources to those areas rather than programs designed to increase motivation. Further, they should reduce the time they spend on product development activities. Finally, American managers should try to give more time to development activities. American managers should insure that German employees fully understand the difference in suggestions and directions.

References Anderson, J.C. and Gerbing, D.W. (1982), ``Some methods for respecifying measurement models to obtain unidimensional construct measurement'', Journal of Marketing Research, Vol. 19 No. 4, pp. 453-60. Baron, R.M. and Kenny, D.A. (1986), ``The moderatormediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations'', Journal of Personality and Social Psychology, Vol. 51 No. 6, pp. 1173-82. Bond, M.H. et al. (1987), ``Chinese values and the search for culture-free dimensions of culture'', Journal of Cross-Cultural Psychology, Vol. 18, June, pp. 143-64. Brown, S.L. and Eisenhardt, K.M. (1995), ``Product development: past research, present findings, and future directions'', Academy of Management Review, Vol. 20 No. 2, pp. 343-78. Clark, T. (1990), ``International marketing and national character: a review and proposal for an integrative theory'', Journal of Marketing, Vol. 54, October, pp. 66-79. Glunk, U., Wilderom, C. and Ogilvie, R. (1996), ``Finding the key to German-style management'', International Studies of Management and Organization, Vol. 26 No. 3, pp. 93-109. Green, R.T. and Langeard, E. (1975), ``A cross-national comparison of consumer habits and innovator characteristics'', Journal of Marketing, Vol. 39, July, pp. 34-41. Hall, E.T. (1966), The Hidden Dimension, Doubleday, New York, NY. Hofstede, G. (1980), Culture's Consequences, Sage Publications, Newbury Park, CA.

Hofstede, G. (1991), Cultures and Organizations: Software of the Mind, McGraw-Hill, London. Hofstede, G. and Bond, M.H. (1988), ``Hofstede's culture dimensions: an independent validation using Rokeach's value survey'', Journal of Cross-Cultural Psychology, Vol. 15 No. 4, pp. 417-33. Ingelfinger, T. (1995), ``Interkulturelle Competenz als Notwendigkeit der Internationalisierung'', Marktforschung & Management, Vol. 3, pp. 103-6. Joint Economic Committee of Congress (1981), Location of High Technology Firms and Regional Economic Development, US Government Printing Office, Washington, DC, June. Joreskog, K.G. and Sorbom, D. (1993), LISREL 8, A Guide to the Program and Applications, SPSS, Chicago, IL. Keller, E. von. (1987), ``Kulturabhaengigkeit der Fuehrung'', in Kieser, A. et al. (Eds), Handwoerter Buch der Fuehrung, Stuttgart Sp. pp. 1285-94. Linton, R. (1945), The Cultural Background of Personality, Appleton-Century Company, New York, NY. Lynn, G.S., Simpson, J.T. and Souder, W.E. (1997), ``Effects of organizational learning and information processing behaviors on new product success'', Marketing Letters, Vol. 8 No. 1, pp. 33-9. Montoya-Weiss, M.M. and Calantone, R. (1994), ``Determinants of new product performance: a review and meta analysis'', Journal of Product Innovation Management, Vol. 11 No. 5, pp. 397-418. Nakata, C. and Sivakumar, K. (1996), ``National culture and new product development: an integrative review'', Journal of Marketing, Vol. 60, January, pp. 61-72. Sauer, P.L. and Dick, A. (1993), ``Using moderator variables in structural equation models'', Advances in Consumer Research, Vol. 20, pp. 637-40. Song, M.X. and Parry, M.E. (1997), ``The determinants of Japanese new product success'', Journal of Marketing Research, Vol. 34, February, pp. 64-76. Song, M.X., Souder, W.E. and Dyer, B. (1997), ``A causal model of the impact of skills, synergy, and design sensitivity on new product performance'', Journal of Product Innovation Management, Vol. 14 No. 2, pp. 88-101. Tse, D.K., Lee, K., Vertinsky, I. and Wehrung, D.A. (1988), ``Does culture matter? A cross-cultural study of executives' choice, decisiveness, and risk adjustments in international marketing'', Journal of Marketing, Vol. 52, October, pp. 81-95.

207

Concept testing of Internet services Muammer Ozer

The author Muammer Ozer is an Assistant Professor in Innovation Management at the Faculty of Business, City University of Hong Kong, Kowloon, Hong Kong. Keywords Consumer behaviour, Market research, Internet, New product development Abstract Studies conducted both in Europe and the USA have consistently suggested that new product development is vitally important for a company's long-term survival. However, it is a risky and uncertain business. Previous studies have indicated that early new product development activities such as concept testing can be instrumental in reducing the risks and uncertainties associated with new products. This paper explores whether a widely used concept testing method can be used to evaluate new Internet initiatives. It presents an empirical study based on a longitudinal survey that involves both the initial predictions and the actual results for two Internet services and shows that concept testing can be a useful tool for companies to evaluate their new Internet services. The paper also provides managerial and research implications. Electronic access The research register for this journal is available at http://www.emeraldinsight.com/researchregisters The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/1460-1060.htm

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . pp. 208±213 # MCB UP Limited . ISSN 1460-1060 DOI 10.1108/14601060210451162

Past studies have consistently suggested that new product development is vitally important for companies to remain competitive in the global marketplace (Yeoh, 1994; Cozijnsen et al., 2000). An essential aspect of new product development is the ``up-front'' evaluation of new product ideas, more formally ``concept testing''. Previous studies have identified the concept testing stage to be one of the most critical stages of the new product development process in creating successful new products (Cooper, 1996; de Brentani and Ragot, 1996). Numerous empirical studies have also shown that concept testing is widely used by companies both in Europe and the US (Nijssen and Lieshout, 1995; Mahajan and Wind, 1992). At the concept testing stage, companies are interested in identifying promising new product ideas for further consideration and development. In a concept testing study, companies usually ask a number of potential buyers to evaluate a particular product concept that is described in a written or simple visual form. The potential buyers state their intentions to purchase the new product tested, and companies use the results to assess its viability before making major financial and non-financial commitments into its development (Ozer, 1999). An important area where the new product concept testing can be applicable is the Internet. Although the Internet offers great competitive advantages, as we all have witnessed, many Internet initiatives could not go beyond the market introduction stage and subsequently failed. It appears that concept testing can be highly relevant for companies in evaluating their new Internet-based services. Thus, the purpose of this paper is to explore the applicability of new product concept testing to Internet services. The following section briefly describes the method of concept testing and discusses related issues. After that, the paper presents an empirical study, showing the use of concept testing and widely used decision rules in estimating the market acceptance of a number of Internet services. Finally, the paper concludes with results and implications.

New product concept testing Behavioural scientists have considered an individual's behavioural intention to be a 208

Concept testing of Internet services

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 208±213

Muammer Ozer

strong determinant of his/her subsequent behaviour (Fishbein and Ajzen, 1975). Thus, behavioural intention measures have been ubiquitously used in a wide range of industries for concept testing of new products as well as for repurchase assessment of already existing products. For example, they were used for automobiles ( Juster, 1966), durable and non-durable goods ( Jamieson and Bass, 1989), services (Infosino, 1986), recorded music products (Lacher and Mizerski, 1994), counterfeit products (Cordell et al., 1996), catalogue products (White, 1996), environmentally friendly products (Marell et al., 1995) and for real estates (Williams et al., 1995). As an application in non-profit setting, they were also used in investigating the roles of such factors as donor knowledge and perceived risk in intention to donate blood (Allen and Butler, 1993). In a typical concept testing study, respondents read a detailed description of the new product concept at hand and state their intentions to purchase it when it becomes available. In addition to stating their intentions to purchase the product, the respondents also answer numerous background questions pertaining to such factors as the demographics, attitudes, interests and opinions (AIOs) and consumer lifestyles. One of the most popular purchase intention measures is a five-point scale item ( Jamieson and Bass, 1989). It can be in the form of the following measure: . Considering the new product description you just read, please indicate your intention to buy it when it becomes available: ± definitely will buy it; ± probably will buy it; ± might/might not buy it; ± probably will not buy it; ± definitely will not buy it. A challenge associated with the purchase intention measures is how to convert them into predictions. For example, let us assume that we measured the purchase intentions of a group of people and found that 25 percent of the surveyed people definitely would buy a new product; 5 percent probably would buy it; 15 percent might or might not buy it; 15 percent probably would not buy it; and 40 percent definitely would not buy it. Given this distribution, the question is how to come

up with a single number indicating that x percent of the target population will buy this product. One way of converting the intention measures into predictions is to use a decision rule that is based on the relationship between people's intentions to buy other similar products and their actual purchase behaviour. Companies use various decision rules (weights) for different intention levels (i.e. five levels ranging from ``definitely will buy it'' to ``definitely will not buy it'') and estimate the percentage of buyers as the weighted sum of the number of individuals in each level. For example, let i be the intention level, wi be the weight associated with that level, ni be the number of people in level i, and N be the number of people surveyed. Then, the percentage of people who would purchase a product is calculated by using the following equation ( Jamieson and Bass, 1989): P …Purchase† ˆ

5 X iˆ1

ni wi … † N

…1†

where: P (purchase) = estimated percentage of the target population who will buy the product; i = intention level (i.e. i = 1 if the answer is ``definitely will buy'' . . . i = 5 if it is ``definitely will not buy''); wi = weight associated with intention level i; ni = number of individuals in intention level i; N = total number of individuals surveyed. A survey with the experienced users of the behavioural intention measures determined several commonly used decision rules ( Jamieson and Bass, 1989). Table I presents the commonly used decision rules identified in that survey. For example, decision rule I indicates that one should only consider the 100 percent of the proportion of those people who state that they definitely will buy the concept tested as an indication of the market Table I Commonly used decision rules

Intention level Definitely will buy Probably will buy Might/might not buy Probably will not buy Definitely will not buy

209

I (%)

II (%)

100 0 0 0 0

28 0 0 0 0

Decision rule III IV (%) (%) 80 20 0 0 0

96 36 0 0 0

V (%)

VI (%)

70 54 35 24 20

75 25 10 5 2

Concept testing of Internet services

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 208±213

Muammer Ozer

demand. For instance, if 25 percent of the surveyed individuals said that they definitely would buy a new product, then the projected demand based on decision rule I is 25 percent (25%*100%). On the other hand, decision rule IV implies that one should take the sum of the 96 percent of the proportion of those people who state that they definitely will buy it and the 36 percent of the proportion of those people who state that they probably will buy it as a prediction. For example, if 40 percent of the individuals surveyed said that they definitely would buy a new product and 10 percent of the people surveyed said that they probably would buy it, then the estimated demand based on decision rule IV is 42 percent (40%*96% + 10%*36%). Given this background, the following section presents an empirical study based on a longitudinal survey about two online services, generates predictions based on the initial intentions and the different decision rules and compares the predictions to the actual results.

Research methodology Two Internet services were conceptualized and developed for people attending a graduate program in business in the northeastern USA so as to observe and compare their initial intentions and actual behaviour. They were online music and sports services. The music service consisted of six different options. The first was a collection of the home pages of individual artists and groups. The second was a music database, containing a wide range of music information. The third was a calendar of events, providing up-to-date worldwide music information about concerts, festivals, and other musicrelated events. The fourth was the classic rock trivia of the month, giving an opportunity to the users to test their music knowledge each month. The fifth was a list of top 100 albums of the year. The final item was related to the frequently asked music questions and their answers. On the other hand, the sports service integrated various sports information sources into a single menu under six general areas. The first area covered numerous professional sports teams from Major League Baseball (MLB), National Basketball Association (NBA), National Hockey League (NHL), and National Football League (NFL). The second

dealt with college sports. The third included specific games and related TV schedules from different leagues including MLB, NBA, NHL, and NFL. The fourth was about the Olympics and frequently asked questions about the Olympic games. The fifth dealt with skiing and worldwide ski-slopes. The sixth was related to soccer. The last area was about the latest sports news and events. Data collection Data collection consisted of two major tasks and was conducted over a one-year period. The first part involved the measurement of the behavioural intentions (i.e. intention to use a particular service). The second part was about the observation of the actual behaviour (i.e. usage of the service concerned). A total of 214 individuals participated in two surveys. There were 160 and 140 usable questionnaires for the music and sports services, respectively. The surveys introduced the service descriptions to the respondents, measured their intentions to use them and collected various background information. The intentions were measured on a five-point scale ranging from ``definitely will use it'' to ``definitely will not use it''. The services were available on an Internet menu page. In order to track the actual users, a private workstation was used as an Internet server, and access to the services was allowed only to those who had a user name and password. Users could access the menu page, but they needed to enter their user name and password to continue. A computer program automatically tracked the inquiries to the server. Individuals in the target population received a letter and an e-mail message approximately every seven to eight weeks throughout the year. The message described the services and stated that they needed to get a user name and password to use the services. It also outlined three ways of getting a user name and password. First, they could visit the Internet menu and fill up the request form online. Second, they could send an email message to the Webmaster. Third, they could send a written request to the Webmaster. The message also included a step-by-step instruction about how to use the services. Upon a request, individuals were issued a user name and password and were offered additional help in case they needed it.

210

Concept testing of Internet services

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 208±213

Muammer Ozer

Results and analyses

Table III Absolute errors based on the different decision rulesa,

After the initial survey, the respondents' answers were categorized into five major groups as required by the prediction equation (1) (e.g. ``those who said that they definitely will use it'' to ``those who said that they definitely will not use it''). Then, the decision rules in Table I were used to calculate the overall predictions for each service. Table II presents the predictions for each service under the different decision rules. For the music service, the predictions were 16 percent, 5 percent, 18 percent, 24 percent, 42 percent, and 22 percent for decision rules I, II, III, IV, V, and VI, respectively. The predictions for the sports service were 31 percent, 9 percent, 28 percent, 36 percent, 46 percent, and 31 percent for decision rules I, II, III, IV, V, and VI, respectively. As was stated earlier, the second part of the data collection was the observation of the actual usage. A computer program automatically tracked the inquiries about the services throughout the year. It kept records of individuals who used the services and the number of times they used them. As it is the case with other online services, such as America-on-Line, there is an initial trial period. Because an individual could log in and out very briefly to see what the services were all about, and in order to eliminate those who were just trying the services, an individual was considered an actual user of a particular service if he/she used it at least three times. Table II presents the percentages of the target population who actually used the services. More specifically, the results indicated that 10 percent of the target population actually used the music and sports services at least three times during the year, respectively. The predictions based on the different decision rules were compared to the actual usage by using a widely used forecasting performance measure, namely the absolute error (|actual-prediction|) for each service. Table III presents the absolute differences between the predictions based on the different Table II Predictions based on the different decision rules

Service

I (%)

II (%)

Music Sports

16 31

5 9

Decision rule III IV (%) (%) 18 28

24 36

V (%)

VI (%)

Actual usage (%)

42 46

22 31

10 10

Service

I (%)

II (%)

Music Sports

6 21

5 1

Decision rule III IV (%) (%) 8 18

14 26

b

V (%)

VI (%)

32 36

12 21

Notes: ajactual-predictionj; bthe lowest absolute errors are italicized

decision rules and the actual usage for each service. As can be seen from the table, the predictions based on the different decision rules differ from the actual usage. For the music service, the absolute errors were 6 percent, 5 percent, 8 percent, 14 percent, 32 percent and 12 percent under decision rules I, II, III, IV, V, and VI, respectively. For the sports service, they were 21 percent, 1 percent, 18 percent, 26 percent, 36 percent, and 21 percent for decision rules I, II, III, IV, V, and VI, respectively. These results show that decision rule II has the lowest absolute error across the different services and consistently outperforms the other rules.

Conclusion and implications Concept testing is an important element of the formal new product development process. Previous research has consistently shown that up-front and pre-development activities stand out as activities that separate winning and losing new products. For example, 74 percent of the 114 industrial new product failures involved either missing or inadequately carried out initial market assessment activities (Cooper, 1975). The uncertainties associated with a new product project are relatively higher at the concept testing stage than at a later stage of the new product development process. Recent studies have also acknowledged the importance of reducing this uncertainty and provided research studies pertaining to the activities at the ``fuzzy-front-end'' of the new product development (Zhang and Doll, 2001). Concept testing can be used to reduce the fuzziness at the initial stage. Managers widely use concept testing for their early new product decisions for several reasons. For example, it is relatively easy to implement. Furthermore, it only requires survey data and a decision rule. As a result, it can be used at the concept testing stage of the new product development process in

211

Concept testing of Internet services

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 208±213

Muammer Ozer

which actual purchase data may not be available. This is very important because many other traditional methods such as time-series analyses require actual purchase data for future projection. In addition, the test can be easily repeated to track the customer opinions over time and to conduct various sensitivity analyses. However, there is no single best decision rule to predict the market acceptance from the initial behavioural intentions. This study tested widely used decision rules and showed that taking the 28 percent of the ``proportion of those people who state that they definitely will buy a product'' as a prediction gives relatively more accurate estimates for similar Internet services. At least at the concept testing stage, decision rule II gives a reasonably consistent ``ballpark'' estimate of the actual usage and can be useful in making early evaluations, which distinguish promising concepts from the non-promising ones (Urban and Hauser, 1993). However, as the concepts are developed into prototypes and products, more sophisticated techniques should be used. Some of these techniques include multiattribute models (Shocker and Srinivasan, 1979), conjoint models (Green and Srinivasan, 1990), pre-test market models (Urban and Katz, 1983; Shocker and Hall, 1986), test market models (Narasimhan and Sen, 1983), and diffusion models (Mahajan et al., 1990). The linkage between initial intentions and actual behaviour has been an important issue for both managers and researchers. Past research has identified numerous factors that explain the linkage between initial intentions and actual behaviour, including product type, and changes in the concept and the environment. Future studies should also investigate the roles of such factors as respondents' culture, involvement, and product knowledge in the relationship between their initial intentions and actual behaviour.

References Allen, J. and Butler, D.D. (1993), ``Assessing the effects of donor knowledge risk on intention to donate blood'', Journal of Health Care Marketing, Vol. 13, Fall, pp. 26-33. Cooper, R.G. (1975), ``Why new products fail'', Industrial Marketing Management, Vol. 4, pp. 315-26.

Cooper, R.G. (1996), ``Overhauling the new product process'', Industrial Marketing Management, Vol. 25, pp. 465-82. Cordell, V.V., Kieschnick, R.L. Jr and Wongtada, N. (1996), ``Counterfeit purchase intentions: role of lawfulness attitudes and product traits as determinants'', Journal of Business Research, Vol. 35 No. 1, pp. 41-53 Cozijnsen, A.J., Vrakking, W.J. and van Ijzerloo, M. (2000), ``Success and failure of 50 innovation projects in Dutch companies'', European Journal of Innovation Management, Vol. 3 No. 3, pp. 150-9. de Brentani, U. and Ragot, E. (1996), ``Developing new business-to-business professional services: what factors impact performance?'', Industrial Marketing Management, Vol. 25, pp. 517-30. Fishbein, M. and Ajzen, I. (1975), Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research, Addison-Wesley, Reading, MA. Green, P.E. and Srinivasan, V. (1990), ``Conjoint analysis in marketing: new developments with implications for research and practice'', Journal of Marketing, Vol. 54, October, pp. 3-19. Infosino, W. (1986), ``Forecasting new product sales from likelihood of purchase ratings'', Marketing Science, Vol. 5, Fall, pp. 372-84. Jamieson, L.F. and Bass, F.M. (1989), ``Adjusting stated purchase intention measures to predict trial purchase of new products: a comparison of models and methods'', Journal of Marketing Research, Vol. 26, August, pp. 336-45. Juster, F.T. (1966), ``Consumer buying intentions and purchase probability: an experiment in survey design'', Journal of the American Statistical Association, Vol. 61, September, pp. 658-96. Lacher, K.T. and Mizerski, R. (1994), ``An exploratory study of the responses and relationships involved in the evaluation of, and in the intention to purchase new rock music'', Journal of Consumer Research, Vol. 21 No. 2, pp. 366-80. Mahajan, V. and Wind, J. (1992), ``New product models: practice, shortcomings and desired improvements'', Journal of Product Innovation Management, Vol. 9, pp. 128-39. Mahajan, V., Muller, E. and Bass, F.M. (1990), ``New product diffusion models in marketing: a review and direction for future research'', Journal of Marketing, Vol. 54, January, pp. 1-26. Marell, A., Davidson, P. and Garling, T. (1995), ``Environmentally friendly replacement of automobiles'', Journal of Economic Psychology, Vol. 16 No. 3, pp. 513-29. Narasimhan, C. and Sen, S.K. (1983), ``New product models for test market data'', Journal of Marketing, Vol. 47, Winter, pp. 11-24. Nijssen, E.J. and Lieshout, K.F.M. (1995), ``Awareness, use, and effectiveness of models and methods for new product development'', European Journal of Marketing, Vol. 29 No. 10, pp. 27-44. Ozer, M. (1999), ``A survey of new product evaluation models'', Journal of Product Innovation Management, Vol. 16 No. 1, pp. 77-94. Shocker, A.B. and Hall, W.G. (1986), ``Pre-test market models: a critical evaluation'', Journal of Product Innovation Management, Vol. 3, pp. 86-107.

212

Concept testing of Internet services

Muammer Ozer

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 208±213

Shocker, A.B. and Srinivasan, V. (1979), ``Multiattribute approaches for product concept evaluation and generation: a critical review'', Journal of Marketing Research, Vol. 16, May, pp. 159-80. Urban, G.L. and Hauser, J.R. (1993), Design and Marketing of New Products, 2nd ed., Prentice-Hall, Englewood Cliffs, NJ. Urban, G.L. and Katz, G.M. (1983), ``Pre-test market models: validation and managerial implications'', Journal of Marketing Research, Vol. 20, August, pp. 221-34. White, T. (1996), ``Solving the modelling puzzle'', Catalog Age, Vol. 13 No 1, pp. 77-8.

Williams, J.D., Qualls, W.J. and Grier, S.A. (1995), ``Racially exclusive real estate advertising: public policy implications for fair housing practices'', Journal of Public Policy & Marketing, Vol. 14 No. 2, pp. 225-44. Yeoh, P.L. (1994), ``Speed to global markets: an empirical prediction of new product success in the ethical pharmaceutical industry'', European Journal of Marketing, Vol. 28 No. 11, pp. 29-49. Zhang, Q. and Doll, W.J. (2001), ``The fuzzy-front-end and success of new product development: a causal model'', European Journal of Innovation Management, Vol. 4 No. 2, pp. 95-112.

213

Introduction

Formalization and innovation revisited Wynand E.J. Bodewes

The author Wynand E.J. Bodewes is Professor of Entrepreneurship at the Rotterdam School of Management, Erasmus University, Rotterdam, The Netherlands. Keywords Innovation, Organizational development, Measurement, Organizational theory Abstract The many studies into the relationships between formalization and innovation have produced little but inconsistent findings. The conceptual and operational definition of the formalization construct is proposed to be one of the reasons for these inconsistencies. It is argued that aggregate (organization-level) measurements of formalization are inappropriate and should be replaced with department-specific or process-specific measurements. Second, it is argued that formalization has been defined in an inconsistent way. However, it is not just the coexistence of different definitions (and their measurements) that is problematic. The exclusion, or improper inclusion, of rule observation from the conceptualization of formalization appears to be a third fallacy. A revised definition of formalization is advanced as a solution to these problems. This definition may prove to be instrumental in determining the true effect of formalization on organizational innovation. Electronic access The research register for this journal is available at http://www.emeraldinsight.com/researchregisters The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/1460-1060.htm European Journal of Innovation Management Volume 5 . Number 4 . 2002 . pp. 214±223 # MCB UP Limited . ISSN 1460-1060 DOI 10.1108/14601060210451171

The question as to which organizational contexts are most supportive to innovation has been on the agenda of students of innovation for many years. In their seminal study, Burns and Stalker (1994) observed that organic systems appear a more suitable organizational context for innovation than mechanistic systems. Aiken and Hage (1971) also concluded that organic systems facilitate innovation. These findings stimulated a line of enquiry that has been with us ever since. Nevertheless, the empirical support for theoretical assertions that structure is important to innovation is inconclusive. Correlates of structural parameters and innovation are instable (Downs and Mohr, 1976; Fiol, 1996; Wolfe, 1994). Downs and Mohr (1976) attribute this instability to conceptual issues. Aiken et al. (1980), and Wolfe (1994) regard the definition and operationalization of innovation as problematic. Inconsistent definition, operationalization and measuring of innovation may indeed have caused instability of findings. However, in this paper it is argued that innovation research has also been crippled by its conceptualization of the formalization construct. Formalization represents the extent to which codified work processes coordinate and control work processes. It is not surprising that students of innovation have tried to identify the effect of formalization on innovation. Formalization is a core dimension of organization structure (Pugh et al., 1968). Yet, reviewing the literature reveals there is less agreement on the meaning of formalization than one would expect with regard to a construct that is at the core of organization theory. Consequently, various definitions and measurements have been used. It needs little clarification to appreciate the implications of inconsistency in definition and measurement for the accumulation of knowledge. Research findings that seem instable, are often incomparable. Therefore, it is impossible to determine whether the instability is caused by the inconsistency in definition and measurement or by other factors. Damanpour's (1991) meta-analysis found ``nonsignificant associations between innovation and formalization''. However, the studies cited in his meta-analysis employ distinctive definitions and measurements for

214

Formalization and innovation revisited

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Wynand E.J. Bodewes

formalization. As a result, it seems that the studies he compared cannot, in fact, be compared. This paper examines how formalization is commonly defined and measured. An alternative definition is presented and discussed. This definition is proposed to help us overcome one of the problems that innovation research has suffered from: the definition and measurement of formalization.

manufacturing to explore and test substantive differences. Aggregate measurements of formalization appear ill-suited to reveal such differences. Empirical studies that measure formalization at the department or process level could not only show that levels of formalization vary, they may also enable us to find out to what extent the relationships between formalization and innovation are department or process specific.

Formalization as an organizational or departmental property

Formalization as a means to control operations

The first problem that has hampered innovation research has been the practice to measure formalization at the organizational level. Whatever the specific definition that is used for formalization, a decision has to be made at what level of aggregation the relationship between formalization and innovation is to be determined. Our review of the literature suggests that few researchers seem sensitive to the importance of this issue. Most seem to imply that formalization is a property that should be measured at the organizational level. Few acknowledge that the extent of formalization may, and often does, vary across departments. This may result from inherent differences in work processes throughout an organization that demand dissimilar levels of formalization. However, the literature suggests that few seem to wonder if or acknowledge that processes in an organization need similar or dissimilar levels of formalization. Essential to many definitions of formalization is that formalization has to do with codification of behavior. Hence, it seems obvious that one has to decide for which behavior one is going to measure the extent of codification. The well known, and widely used, measuring instrument for formalization developed by the Aston group (Pugh et al., 1968) and Hall (1963) provides an aggregate measure at the organizational level. The ``type of behavior'' that is codified, e.g. developing a new product versus assembling an existing product, remains unspecified. More insight might result from studies that are more explicit on the kind of behaviors that were codified. If one seeks to understand the way formalization affects organizational innovation, one may want to include a measurement of formalization in R&D and another in

Levels of measurement and aggregation create an important problem to students of innovation. It is not the only problem that they will encounter. Also, the inconsistencies in definition pose a challenge. A starting point to understanding the meaning of formalization is to show how other researchers have defined and used this construct. From this overview it follows that researchers have different views regarding the meaning and measurement of formalization. This has serious implications for what is being studied. Our literature review suggests few researchers are sensitive to the importance of careful definition. Pennings (1973) and Sathe (1978) pointed to differences in measuring. This paper goes further by pointing to differences in definitions. Formalization is ``integral to the concept of organization'' (Etzioni, 1975, p. 457). Consequently, researchers have given formalization a prominent position in their conceptual frameworks and theories of organization. Formalization has been presented as a primary dimension of organization structure (Pugh et al., 1963), a design parameter of formal structure (Mintzberg, 1979), and as a control mechanism (Collins et al., 1988). The construct is not only prominent within organization theory. It also features in related fields such as technology management, strategy, organizational behavior and organizational communication. Most definitions of formalization seem to assume that this construct can best be conceptualized by its function in organizational coordination and control. Although formalization can have multiple functions (Walsh and Dewar, 1987), its role in organizational control is most salient.

215

Formalization and innovation revisited

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Wynand E.J. Bodewes

Control theory is, therefore, a suitable starting point to define formalization. Simon (1976, p. 147) defines formal organization as: ``a set of abstract, more or less permanent relations that govern the behavior of each participant''. Organizations are ``systems of coordinated action among individuals and groups'' (March and Simon, 1993, p. 2). Several mechanisms can be used to coordinate and control the behavior of social collectives and their members. A ``mechanism'' which all social collectives can apply to handle problems of coordination and control is to reduce or eliminate the need for social coordination and control (Merchant, 1985; Den Hertog and Wielinga, 1992). One way to achieve this is to prevent control problems from having a significant impact on the organization, that is, to minimize the potential losses that arise when such problems occur (Merchant, 1985). Organizations can also apply controls that are ``embedded in the physical structure'' (Edwards, 1979, p. 20) to eliminate or reduce the need for social control. Ouchi (1979, 1980) describes three additional mechanisms through which social collectives can coordinate and control their operations and outputs. Frictionless markets theoretically provide the best coordination mechanism for all (organizational) processes. However, market imperfections can, and often do, render other mechanisms more efficient than the market mechanism (Williamson, 1975). Ouchi (1979) presents bureaucratic control and clan control as the two other mechanisms that can coordinate and control social behavior. Socialization processes provide the underpinning for clan control. Clan control operates through shared norms and values. Ouchi reserves the label ``clan control'' for socialization processes within an organization. Socialization in other collectives than organizations is given different labels. In Ouchi's (1979) terminology, ``cultural control'' refers to socialization at the level of a political unit (e.g. a country), and ``professional control'' refers to the socialization of social actors that share a set of norms and values, but who occupy different organizations. Shared norms and values (informal rules) that coordinate (direct) behavior, and comparison of actual behavior to these informal rules (behavior monitoring) is the essence of controls that rest on socialization processes. The informal rules

provide the standards required for (cybernetic) control. Formal and informal rules act as coordination devices by ``providing a memory for handling routine situations'', because they ``eliminate the need for treating each situation as new'' which reduces the amount of communication and decision making required (Galbraith, 1973, p. 10). Rules not only coordinate, they can also transmit past learning and ``control (make predictable) behavior within the firm'' (Cyert and March, 1992, p. 124). Formal rules and programs are also sources of internal (Walsh and Dewar, 1987) and external (Sitkin and Bies, 1993) legitimization.

Formalization as codified structural controls Formal and informal rules provide standards for behavior or outcomes. Behavior controls operate through rules that outline the conduct of the collective, a department or an individual actor. A recipe is a good example of a rule (or set of rules) that provides a behavioral template to a social actor. Such rules specify the behavior that supposedly ensures that the outcome(s) of this behavior will be acceptable. Output controls do not specify the ``production process'', they specify parameters and criteria that signal the acceptability of the output. Technical standards, that define nuts and bolts, can be used in quality control to evaluate the acceptability of a process outcome (e.g. a batch of bolts), and to take corrective action if needed. Formalization provides the explicit rules that are required for bureaucratic control. Walsh and Dewar (1987) argue that formalization has three functions. Formalization provides a shared language that facilitates efficient communication. It can code complex behavior or outcomes to simple rules. For a formal rule to be efficient, it is not always necessary to include a detailed description. This is especially true if a set of formal rules is backed up by a set of informal rules that enables the correct interpretation of the formal rules' meaning. If such (informal) interpretation rules exist, then a formal rule can be extremely condensed indeed. Blueprints such as those used in construction, or process flow diagrams used in chemical

216

Formalization and innovation revisited

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Wynand E.J. Bodewes

industries are good examples of condensed descriptions that have little meaning if one does not have the knowledge required to interpret them. Hence, a risk of coding is that the message is misinterpreted. However, one should not forget that to an extent all communication is coded. The more a message is condensed, the more important it becomes for the receiver of that message to have the knowledge that is required to interpret the message correctly. Formalization not only embodies codified knowledge, it also defines appropriate and inappropriate behavior (or outcomes) (Walsh and Dewar, 1987). Formalization signals when an output or behavior will be unacceptable. Formalization explicates degrees of freedom. Formalization's ability to serve as a standard of distributive justice is the last function that Walsh and Dewar (1987) present. The existence of a formal rule legitimizes rule observation and enforcement of compliance. Formalization plays a central role in all bureaucratic control, not only by providing a standard for what is to be done, also for what is not to be done, which in turn legitimizes rule enforcement. Formalization provides explicit procedural standards for acceptable as well as unacceptable behaviors and outcomes. This description provides several components for an initial definition. A first definition of formalization is: D1. Formalization is the extent to which procedures for behavior and outputs are documented. Reviewing the literature for definitions of formalization, we see that this type of definition is commonly used. Our survey of the literature was not exhaustive. We searched for empirical papers that report formalizationinnovation relationships. Table I lists several of the definitions that were found in the literature that resemble the above definition. This definition of formalization is quite common. Many adopt similar definitions. The Aston group (Pugh et al., 1963, 1968, 1969a, 1969b; Pugh, 1973; Inkson et al., 1970) has been most influential, not only in defining formalization, but also in suggesting and developing ways to measure other characteristics of organization structure. The Aston studies employed ``institutional'' measures by relying on documents and official informants (Pennings, 1973).

A similar approach has been used by many researchers who developed measuring scales for formalization that count the number of documents of a set list that is available in the organization. The Aston group started measuring formalization with a list of 38 documents ``each of which could be used by any known work organization'' (Pugh et al., 1968). This list has been reduced to measuring instruments with ten (Zmud, 1982), nine (Inkson et al., 1970; Blau et al., 1976; Kerr and Jermier, 1978) and eight (House and Rizzo, 1972) items. After reviewing the relevant literature, it can be concluded that this type of definition and measuring is widely used. Defining formalization as the existence of documented procedures indicates it is one of the elements required for bureaucratic control. Comparison of behavior/output to the standard, and subsequent corrective action is also required for bureaucratic control to exist. Therefore, this definition of formalization by itself does not allow us to regard formalization as a coordination/control mechanism. More is needed for documented procedures to coordinate and control. We would need to know if the documented procedures are actually being used, e.g. to compare processes/outcomes with the documented procedures, and/or to instigate corrective actions. The ``Aston'' type definition and its measurements, therefore, do not describe the way formalization is being used in control. Measurements that fit the first type of definition provide an indication of the extent to which an organization has filled its bookshelves with standard operating procedures, organizational charts, work instructions and the like. They represent espoused controls. These may differ from actual control through enactment of codified rules and procedures.

Formalization as enacted codified control The exclusion of the use that is being made of these documents is perhaps not problematic if one is only interested in the amount of paperwork that exists. However, such a definition is unfit for research designs that seek to determine the effect of formalization on processes, behavior and outcomes at the individual, group, department or organization

217

Formalization and innovation revisited

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Wynand E.J. Bodewes

Table I Formalization defined as codified procedures Exemplary definitions

Similar definitions used by:

How far communications and procedures are written down and filed (Pugh et al., 1963)

Pugh et al., 1968; Inkson et al., 1970; Mansfield, 1972; House and Rizzo, 1972; Child, 1973; Blau et al., 1976; Kerr and Jermier, 1978; Blau and McKinley, 1979; Organ and Greene, 1981; Zmud, 1982; Ettlie et al., 1984; Podsakoff et al., 1986; Miller and DroÈge, 1986; Miller, 1987; Adler and Borys, 1996

[T]he degree to which a codified body of rules, procedures or behavior prescriptions is developed to handle decisions and work processes (Pierce and Delbeq, 1977) The degree to which divisions of labor and procedures are explicit (e.g. written or espoused in the form of guidelines) rather than implicit (Glisson and Martin, 1980). The control of job activities by administrative rules and procedures. The extent to which standard practices, policies and position responsibilities have been explicitly formalized by the organization (Organ and Greene, 1981) Control by plan (Hull and Hage, 1982) The degree to which the norms of an organization are explicitly formulated (Price and Mueller, 1986)

level. Such studies should explicitly take into account the extent to which documented procedures are being used. Reviewing the literature revealed that this is often omitted. For example, Organ and Greene (1981) use the measurement scale developed by House and Rizzo (1972) to determine the effect of formalization on characteristics such as role ambiguity, organizational identification, role conflict and alienation. Organ and Greene (1981) define formalization as ``the control of job activities by administrative rules and procedures''. One would expect that their theoretical model includes a construct like rule observation. Neither their theoretical model, nor the instruments that they use to test this model, include such a construct. Also, a subsequent replication of the Organ and Greene study conducted by Podsakoff et al. (1986) does not truly operationalize formalization as a combination of the extent to which documented rules exist, and the extent to which the rules are being observed. One wonders how formalization could influence behavior and outcomes when documented rules are not being observed. There could be an indirect effect, caused by resources being depleted in creating documented rules. Such depleted resources

can put a cap on any effort to innovate. This would result in an indirect negative effect. A positive effect could result from investment of surplus revenues generated by innovating products or processes in more extensive codification. However, we propose that these indirect effects provide theoretically weak explanations for relationships between formalization and behavior (e.g. new product development) or between formalization and behavioral outcomes (e.g. new products). One would expect that the importance of rule observation is taken into account when a causal relationship between codification and behavior is theorized. Unfortunately, researches adopting definitions similar to the ones in Table I do not acknowledge that codified rules need to be (implicitly) referenced and enacted. Their practice to define and operationalize formalization as the extent to which documented procedures exist appears to leave out an important element. The use that is made of the documented procedures should be included to establish the meaning of formalization as a structural parameter and its meaning as a control mechanism. This implies that rule observation is to be included in the definition of formalization.

218

Formalization and innovation revisited

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Wynand E.J. Bodewes

Fortunately, our review suggests that several researchers include rule observation in their definitions of formalization. For example, Walsh and Dewar (1987) note that ``little used or long-forgotten [documented] regulations can hardly be considered `formal' since few, if any, actors know that they exist or remember them''. They argue for the inclusion of a behavioral component in the definition of formalization. Such definitions provide a better starting point to determine and explain relationships between formalization and alienation, or between formalization and innovation, than a definition that only includes the existence of documented procedures. Based on our analysis of the definitions that were identified in the literature we propose a definition that incorporates enactment: D2. Formalization is the extent to which documented procedures that communicate desired outputs and behavior exist and are being used. Table II lists some of the definitions that we identified in the literature that resemble the above conceptualization: The two tables should not be used to assess which of the two definitions has more followers. Our aim is to show that both have been used in influential studies.

Measuring the extent of formalization The difference between both definitions becomes even more profound when their results are compared. The first group of

definitions tends to see the existence of a number of documented standards from a closed list as a proxy for the extent of formalization. Hence, firms with comparatively many standard operating procedures are regarded as more extensively formalized. We believe such measurements provide flawed proxies for the formalization construct that we feel should be conceptualized. The extent to which the codified procedures are being enacted should be included in establishing the level of formalization. The second group of formalization definitions adds rule observation to the notion of codification as an additional dimension. The instrument developed by Hall (1963) is often referenced and used. This instrument seeks to provide a direct measurement through items such as: We are to follow strict operating procedures at all times (Hall, 1963).

Note that this item does not refer to actual observation but to espoused observation. Others constructed measuring instruments that deploy sets of indicators for both codification and rule observation. Hage and Aiken (1967b) were among the first to develop such an instrument. In their case, codification represents the degree of work standardization while observation represents the latitude of behavior that is tolerated from standards (Hage and Aiken, 1967b). Hall's (1963) instrument portrays formalization as an ``espoused'' (Argyris, 1992) parameter of organization structure whereas Hage and Aiken (1967b) portray formalization as a parameter of organization structure in action.

Table II Formalization defined as enacted codified procedures Exemplary definitions

Similar definitions used by:

Hall, 1963; Aiken and Hage, 1971; Hage and Dewar, The degree of work standardization and the 1973; Glisson and Martin, 1980; Gaines and Jermier, amount of deviation that is allowed from 1983; Walsh and Dewar, 1987 standards (Aiken and Hage, 1966) Represents the use of rules in an organization. [It has] two indicators: job codification and rule observation (Hage and Aiken, 1967b) [T]he emphasis placed within the organization on following specific rules and procedures in performing one's job (Zaltman et al., 1984, p. 138) The rules and procedures designed to handle the contingencies faced by organizational members [ . . . ] there is a conceptual overlap between job formalization and the degree to which standardized procedures are followed in performing a task (Malik and Wilson, 1995)

219

Formalization and innovation revisited

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Wynand E.J. Bodewes

Sathe (1978) has made a similar distinction between designed and emergent organizational structure. Our suggestion that we should conceptualize formalization as a structural parameter in use renders the second group of definitions more suitable to determining and explaining the relationships between formalization and such constructs as alienation or innovation. One would hope that our survey of the literature shows that only measuring instruments that fit the second group of definitions have been used to test formalization-innovation relationships. To our surprise, this was not the case. Illustrative is that of the 11 studies that tested the existence of a relationship between formalization and organizational innovation that were included in Damanpour's metaanalysis (1991), four studies (Blau and McKinley, 1979; Zmud, 1982; Hull and Hage, 1982; Ettlie et al., 1984) treat formalization as codification, five studies (Hage and Aiken, 1967a; Aiken and Hage, 1971; Kaluzny et al., 1974; Kim, 1980; Cohn and Turyn, 1980) treat it as codification and rule observation, whereas two studies (Corwin, 1975; Aiken et al., 1980) measured task standardization. Task standardization was used by Damanpour as a proxy for formalization. The substantive differences between these measurements should make any attempt at meta-analysis menacing. Attempts to generalize from such studies can only suggest that findings appear to be instable. The coexistence of the two groups of definitions is problematic. It renders outright comparisons of empirical findings deceptive. Hence, studies that exclude observation from their definitions and measurements should be interpreted with proper caution. The second group is preferable to the counting of documents. Rule observation must be included in the definition of formalization when the relation between formalization and behavior (and behavioral outcomes) is to be explained. A consequence of our argumentation is that organization theorists should be less tolerant to studies that employ definitions and measurements of formalization that do not incorporate rule observation. However, it seems insufficient to define and measure formalization as a two-dimensional construct.

Formalization as interaction of codification and observation The issue that remains, is how documentation and rule observation should be aggregated to conceptualize formalization as a system of enacted codified procedures. Researchers that see formalization as a two-dimensional construct refrain from discussing in what way the two dimensions relate to formalization. Aiken and Hage (1966) note that: . . . [a] high degree of formalization implies not only a preponderance of rules defining jobs and specifying what is to be done, but also the enforcement of those rules.

Hence, they acknowledge job codification and rule observation as the two dimensions that embody the formalization construct. Aiken and Hage (1966) measured formalization through a set of five items for job codification, and two items for rule observation. Each dimension's index was computed by averaging its items. The two aspects were subsequently used as independent proxies for the extent of formalization. Our survey of the literature shows that this approach is used by all researchers who define formalization as the coexistence of documented standards and rule observation. We feel that this is equally distressing as the exclusion of rule observation from a definition of formalization. One would hope that researchers explicate why and how the two dimensions are combined. Surveying the literature indicated that most researchers ignore this important issue. The dimensions are commonly used as independent proxies of the underlying construct. When indicators can compensate for each other this would be a valid approach. However, when it is the interaction of two orthogonal dimensions that embodies a construct, dimensions themselves cannot be used as proxies for the construct. Formalization seems such a construct that should be represented as the joint effect of two unrelated dimensions. This idea can be illustrated by a simple example. Let us assume that codification is measured through the existence of a rule manual and a documented organization chart (0 = no manual and no organization chart, 1= the existence of either a manual or an organization chart, and 2 = the existence of a manual as well as an organization chart), and rule observation is measured through the reference that is made to the documented

220

Formalization and innovation revisited

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Wynand E.J. Bodewes

standards (0 = never, 1 = occasionally, and 2 = frequently). Most readers will agree that frequent use of both the rule manual and the organization chart results in the highest degree of formalization. This would be in line with what the second group of definitions appears to aim for. Walsh and Dewar (1987), for example, state that: . . . formalization refers more to what the rules do, [ . . . ] rather than their number.

It is, however, uncommon to specify how codification and observation can represent what the rules do. We propose that it is the joint effect of codification and observation that matters, not their individual levels. This joint effect of codification and rule observation conceptualizes formalization as a parameter of organization structure in use. The common way to model such a construct would be to model it as a moderated effect. Observation would then be the moderating dimension and codification the moderated dimension. It is the interaction term of codification and observation that captures the essential meaning of formalization. The definition of formalization and its measurements must expose that extensive codification without any significant rule observation, or extensive rule observation without any significant codification, has a lesser impact (on behavior or outcomes) than a situation where moderate codification is combined with moderate rule observation. Hence, it is not important to know just the extent of codification or the extent of observation to get a feeling for the extent of formalization. One needs to know both. The third definition for formalization does that: D3. Formalization is the extent to which documented standards are used to control social actors' behavior and outputs. Such a definition is felt to represent the meaning of this construct better than the two definitions that were discussed above as representatives of common definitions. It would seem that students of innovation should not neglect the definition of their constructs. Also the measurement of formalization deserves more attention. This may ultimately contribute to consistency in terminology. Consistency will not result from mimicking definitions but it may result from careful conceptualization and measuring of our constructs.

Conclusion This paper discusses three problems with previous studies that relate formalization to organizational innovation. First, it seems that the level of analysis is more important than is usually acknowledged. One may doubt whether one should use aggregated measurements for formalization at the organization level when substantial differences in the extent of formalization across departments are to be expected. We argue that it is to be prudent to better specify the kinds of behavior for which the extent of formalization is to be defined and measured. Comparative analysis of department or taskspecific measures may help us to understand how, when, and where formalization has different effects on organizational innovation. The two other problems relate to the definition and measurement of formalization. Reviewing the literature identified two distinctive groups of definitions. In addition, these definitions fail to provide convincing results. It seems plausible that this may have contributed to the instability of findings. A revised definition that conceptualizes formalization as the coexistence of job codification and rule observation, combined with a view of formalization as the joint effect of codification and observation is proposed to overcome one of the problems that has crippled previous research. Future studies that seek to test how formalization affects innovation may therefore want to discuss and consider the way they define and measure formalization more extensively. It seems a bit too hasty to conclude that: . . . merely pushing harder to specify organizational level moderators is not likely to solve our problem of accumulating masses of inconsistent research results (Fiol, 1996, p. 1013).

If inconsistencies result from unsatisfactory definitions then we should start by examining the way we define and measure our constructs and variables.

References Adler, P.S. and Borys, B. (1996), ``Two types of bureaucracy: enabling and coercive'', Administrative Science Quarterly, Vol. 41 No. 1, pp. 61-89. Aiken, M. and Hage, J. (1966), ``Organizational alienation: a comparative analysis'', American Sociological Review, Vol. 31 No. 4, pp. 497-507.

221

Formalization and innovation revisited

Wynand E.J. Bodewes

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Aiken, M. and Hage, J. (1971), ``The organic organization and innovation'', Sociology, Vol. 5 No. 1, pp. 63-82. Aiken, M., Bacharach, S.B. and French, J.L. (1980), ``Organizational structure, work process, and proposal making in administrative bureaucracies'', Academy of Management Journal, Vol. 23 No. 4, pp. 631-52. Argyris, C. (1992), On Organizational Learning, Basil Blackwell, Oxford. Blau, J.R. and McKinley, W. (1979), ``Ideas, complexity, and innovation'', Administrative Science Quarterly, Vol. 24 No. 2, pp. 200-19. Blau, P.M., Falbe, C.M., McKinley, W. and Tracy, P.K. (1976), ``Technology and organization in manufacturing'', Administrative Science Quarterly, Vol. 21 No. 1, pp. 20-40. Burns, T. and Stalker, G.M. (1994), The Management of Innovation, Oxford University Press, Oxford. Child, J. (1973), ``Predicting and understanding organization structure'', Administrative Science Quarterly, Vol. 18 No. 2, pp. 168-85. Cohn, S.F. and Turyn, R.M. (1980), ``The structure of the firm and the adoption of process innovations'', IEEE Transactions on Engineering Management, Vol. 27 No. 4, pp. 98-102. Collins, P.D., Hage, J. and Hull, F.M. (1988), ``Organizational and technological predictors of change in automaticity'', Academy of Management Journal, Vol. 31 No. 3, pp. 512-43. Corwin, R.G. (1975), ``Innovation in organizations: the case of schools'', Sociology of Education, Vol. 48 No. 1, pp. 1-37. Cyert, R.M. and March, J.G. (1992), A Behavioral Theory of the Firm, Blackwell Business, Cambridge, MA. Damanpour, F. (1991), ``Organizational innovation: a meta-analysis of effects of determinants and moderators'', Academy of Management Journal, Vol. 34 No. 3, pp. 555-90. Den Hertog, J.F. and Wielinga, C. (1992), ``Control systems in dissonance: the computer as an ink blot'', Accounting, Organisations and Society, Vol. 17 No. 2, pp. 103-27. Downs, G.W. and Mohr, L.B. (1976), ``Conceptual issues in the study of innovation'', Administrative Science Quarterly, Vol. 21 No. 4, pp. 700-14. Edwards, R.C. (1979), Contested Terrain: The Transformation of the Workplace in the Twentieth Century, Basic Books, New York, NY. Ettlie, J.E., Bridges, W.P. and O'Keefe, R.D. (1984), ``Organizational strategy and structural differences for radical versus incremental innovation'', Management Science, Vol. 30 No. 6, pp. 682-95. Etzioni, A. (1975), A Comparative Analysis of Complex Organizations, The Free Press, New York, NY. Fiol, C.M. (1996), ``Squeezing harder doesn't always work: continuing the search for consistency in innovation research'', Academy of Management Review, Vol. 21 No. 4, pp. 1012-21. Gaines, J. and Jermier, J.M. (1983), ``Emotional exhaustion in a high stress organization'', Academy of Management Journal, Vol. 26 No. 4, pp. 567-86. Galbraith, J.R. (1973), Designing Complex Organizations, Addison-Wesley, Reading, MA. Glisson, C.A. and Martin, P.Y. (1980), ``Productivity and efficiency in human service organizations as related

to structure, size, and age'', Academy of Management Journal, Vol. 23 No. 1, pp. 21-37. Hage, J. and Aiken, M. (1967a), ``Program change and organizational properties: a comparative analysis'', American Journal of Sociology, Vol. 73 No. 1, pp. 503-19. Hage, J. and Aiken, M. (1967b), ``Relationship of centralization to other structural properties'', Administrative Science Quarterly, Vol. 12 No. 1, pp. 72-92. Hage, J. and Dewar, R. (1973), ``Elite values versus organizational structure in predicting innovation'', Administrative Science Quarterly, Vol. 18 No. 3, pp. 279-90. Hall, R.H. (1963), ``The concept of bureaucracy'', American Journal of Sociology, Vol. 69 No. 1, pp. 32-40. House, R.J. and Rizzo, J.R. (1972), ``Toward the measurement of organizational practices: scale development and validation'', Journal of Applied Psychology, Vol. 56 No. 5, pp. 388-96. Hull, F. and Hage, J. (1982), ``Organizing for innovation: beyond Burns and Stalker's organic type'', Sociology, Vol. 16 No. 4, pp. 564-77. Inkson, J.H., Pugh, D.S. and Hickson, D.J. (1970), ``Organization content and structure: an abbreviated replication'', Administrative Science Quarterly, Vol. 15 No. 3, pp. 318-29. Kaluzny, A.D., Veney, J.E. and Gentry, J.T. (1974), ``Innovation of health services: a comparative study of hospitals and health departments'', Health and Society, Vol. 52 No. Winter, pp. 51-82. Kerr, S.J. and Jermier, J.M. (1978), ``Substitutes for leadership: their meaning and measurement'', Organizational Behavior and Human Performance, Vol. 22 No. 3, pp. 375-403. Kim, L. (1980), ``Organizational innovation and structure'', Journal of Business Research, Vol. 8 No. 2, pp. 225-45. Malik, S.D. and Wilson, D.O. (1995), ``Factors influencing engineer's perceptions of organizational support for innovation'', Journal of Engineering Technology Management, Vol. 12 No. 3, pp. 201-18. Mansfield, R. (1972), ``Bureaucracy and centralization: an examination of organizational structure'', Administrative Science Quarterly, Vol. 18 No. 4, pp. 477-88. March, J.G. and Simon, H.A. (1993), Organizations, Blackwell Business, Cambridge, MA. Merchant, K.A. (1985), Control in Business Organizations, Ballinger, Cambridge, MA. Miller, D. (1987), ``Strategy making and structure: analysis and implications for performance'', Academy of Management Journal, Vol. 30 No. 1, pp. 7-32. Miller, D. and DroÈge, C. (1986), ``Psychological and traditional determinants of structure'', Administrative Science Quarterly, Vol. 31 No. 4, pp. 539-60. Mintzberg, H. (1979), The Structuring of Organizations: a Synthesis of the Research, Prentice-Hall, Englewood Cliffs, NJ. Organ, D.W. and Greene, C.N. (1981), ``The effects of formalization on professional involvement: a compensatory process approach'', Administrative Science Quarterly, Vol. 26 No. 2, pp. 237-52.

222

Formalization and innovation revisited

Wynand E.J. Bodewes

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 214±223

Ouchi, W.G. (1979), ``A conceptual framework for the design of organizational control mechanisms'', Management Science, Vol. 25 No. 9, pp. 833-48. Ouchi, W.G. (1980), ``Markets, bureaucracies, and clans'', Administrative Science Quarterly, Vol. 25 No. 1, pp. 129-41. Pennings, J.M. (1973), ``Measures of organizational structure: a methodological note'', American Journal of Sociology, Vol. 75 No. 3, pp. 686-704. Pierce, J.L. and Delbeq, A.L. (1977), ``Organization structure, individual attitudes, and innovation'', Academy of Management Review, Vol. 2 No. 1, pp. 27-37. Podsakoff, P.M., Williams, L.J. and Todor, W.D. (1986), ``Effects of organizational formalization on alienation among professionals and nonprofessionals'', Academy of Management Journal, Vol. 29 No. 4, pp. 820-31. Price, J.L. and Mueller, C.W. (1986), Handbook of Organizational Measurement, Pitman Publishing Inc., Marshfield, MA. Pugh, D.S. (1973), ``The measurement of organization structures: does context determine form?'', Organisational Dynamics, Vol. 1 No. 1, pp. 19-34. Pugh, D.S., Hickson, D.J. and Hinings, C.R. (1969a), ``An empirical taxonomy of structures of work organizations'', Administrative Science Quarterly, Vol. 14 No. 1, pp. 115-26. Pugh, D.S., Hickson, D.J., Hinings, C.R. and Turner, C. (1968), ``Dimensions of organization structures'', Administrative Science Quarterly, Vol. 13 No. 1, pp. 65-105. Pugh, D.S., Hickson, D.J., Hinings, C.R. and Turner, C. (1969b), ``The context of organization structures'',

Administrative Science Quarterly, Vol. 14 No. 1, pp. 91-114. Pugh, D.S., Hickson, D.J., Hinings, C.R., Macdonald, K.M., Turner, C. and Lupton, T. (1963), ``A conceptual scheme for organizational analysis'', Administrative Science Quarterly, Vol. 8 No. 3, pp. 289-315. Sathe, V. (1978), ``Institutional versus questionnaire measures of organizational structure'', Academy of Management Journal, Vol. 21 No. 2, pp. 227-38. Simon, H.A. (1976), Administrative Behavior: A Study of Decision-making Processes in Administrative Organization, The Free Press, New York, NY. Sitkin, S.B. and Bies, R.J. (1993), ``The legalistic organization: definitions, dimensions, and dilemmas'', Organization Science, Vol. 4 No. 3, pp. 345-51. Walsh, J.P. and Dewar, R.D. (1987), ``Formalization and the organizational life cycle'', The Journal of Management Studies, Vol. 24 No. 3, pp. 215-31. Williamson, O.E. (1975), Markets and Hierarchies: Analysis and Antitrust Implications, A Study in the Study of Internal Organization, The Free Press, New York, NY. Wolfe, R.A. (1994), ``Organizational innovation: review, critique, and suggested research directions'', Journal of Management Studies, Vol. 31 No. 3, pp. 405-31. Zaltman, G., Duncan, R. and Holbeck, J. (1984), Innovations and Organizations, John Wiley & Sons, New York, NY. Zmud, R.W. (1982), ``Diffusion of modern software practices: influence of centralization and formalization'', Management Science, Vol. 28 No. 12, pp. 1421-31.

223

Open or closed strategy in developing new products? A case study of industrial NPD in SMEs Martti Tapio Lindman

The author Martti Tapio Lindman is Principal Lecturer at Vaasa Polytechnic, Vaasa, Finland. Keywords New product development, Small- to medium-sized enterprises, Strategy, Product innovation Abstract This case study reports the quality of industrial new product development in five small- to medium-sized enterprises (SMEs) in the Finnish metal industry. The findings indicate that SMEs tend to lack a long run perspective; that the role of new products in business strategy calls for clarity and that the whole goal setting as to future new product efforts is limited. The fit between market requirements and firms' own resources is managed due to the flexibility SMEs have and by relying on an in-house knowledge base generated through a close understanding of user conditions. In this respect SMEs are apt to rely on reactive and closed new product strategies only. Even if successful in the past, such strategies risk being unable to identify and take advantage of any business opportunities outside the present product scope. Also, the increasing need of networking may turn out a threat if SMEs cannot establish more open development strategies. Electronic access The research register for this journal is available at http://www.emeraldinsight.com/researchregisters The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/1460-1060.htm European Journal of Innovation Management Volume 5 . Number 4 . 2002 . pp. 224±236 # MCB UP Limited . ISSN 1460-1060 DOI 10.1108/14601060210451180

Introduction The literature regarding industrial new product development (NPD) indicates that differences in firm-specific NPD behaviour occur because industrial firms differ in their size, organizational specialization and availability of resources. Large firms especially are expected to comprise the major source of technology and innovations (Pavitt, 1991), since the lack of relevant skills and resources is considered a potential weakness for innovation in small- to medium-sized enterprises (SMEs) (e.g. Freel, 1998). Rothwell instead states that neither small nor large companies possess any extraordinary advantages over each other in innovation and that success is people embodied rather than resource embodied (Rothwell, 1992, 1983). This implies that an industrial firm's ability to take advantage of emerging new product opportunities is largely a matter of management skills available and of the corresponding ability to create and apply some new knowledge, which innovation, by definition, calls for. When industrial NPD is looked at from this point of view ± how to build and exploit new knowledge ± firms face the dilemma of management orientation in the sense that any useful new knowledge, applicable in terms of NPD, can be searched and utilized in a number of different ways. Analogically, the same problem as in purchasing, to make or buy, is in question. The acknowledged dilemma can be approached by following either a more open or closed innovation strategy (NystroÈm, 1990). In the first case, a more open strategy means that firms seek cooperation and flexibility by utilizing any knowledge they can make available from external sources instead of resorting solely to their own resources. On the contrary, a more closed strategy means that, instead of coming from cooperation, the new knowledge required is generated by utilizing internal resources only ± even at the cost of lesser flexibility. Following a more open strategy further implies that one can basically learn from others. In the contrary case, the people in the firm believe they know best what to do themselves. Whichever the case, the inherent new product competition means that there are both winning and losing firms, thus impacting the prevailing economic situation and employment opportunities in a nearby and

224

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

related environment (e.g. Christensen, 1997). Given the demands of competition and any growth expectations related to SMEs, the new product competences, which can be met at the level of SMEs, turn out to be the key to survival in the long run. Understanding the SMEs' NPD behaviour from this point of view becomes a prerequisite for developing and implementing working public economic and technological policies. In this respect, the purpose of this study is to outline the present quality of NPD in SMEs of the Finnish metal industry, by identifying and assessing the level of strategic management of new products according to which NPD is actually managed and implemented. The need for openness, which the increasing trend of augmenting interfirm networking implies (e.g. Zineldin and Jonsson, 2000; Bengtsson and Kock, 1999), suggests that the importance of cooperation regarding NPD is increasing also and that SMEs can benefit through a proper cooperation. From this point of view, the target of this study is to evaluate whether a more open or closed innovation strategy, (as it is used synonymously here, a NPD strategy), is searched for in general.

Methodology The study is based on a comparative case research setting consisting of five case firms. Given that a highly complicated and multidimensional phenomenon is in question, the study employs a holistic approach and rests on the theory of strategic management of new products and the relevant new product performance literature available. Those new product performance factors, which have proved to be of importance both from the strategy and implementation point of view have been especially examined to establish the evaluation of NPD. The list of the literature utilized and the corresponding factors extruded is in Appendix 1 (Table AI). The principle of benchmarking has been applied to the case firms in the sense that any parallels are searched for between the established evaluation base of new product performance and corresponding indications to which the data collected by in-depth interviewing have given reason (c.f. Cooper, 1998). This has made it possible to avoid the risk that holism may cause too high a level of abstraction (Yin, 1989) and to investigate the

new product activity from a number of different angles by which strategic, organizational, cultural or marketing aspects of NPD can be made manageable at the same time. To create a more compact parallel matrix for benchmarking purposes, the performance factors have been categorized by using conventional concepts for further analysis. For this purpose, each performance factor has been further elaborated along a key statement indicating if the terms of performance are met. The terms of performance, as suggested by the literature, are listed in Appendix 2 (Table AII). The data collection has taken place during the summer of 2000 through personal in-depth interviews of relevant managers in charge. For a list of persons interviewed and their professional position, see Appendix 3. The interviewing has taken place through open-ended questions, which have been made along the themes that each performance factor represents. The questioning was completed up to the point where the researcher was able to make as credible a judgement as possible if the terms of performance were met or not in an observable way. For a more detailed analysis, all interviews have been recorded and typewritten. To avoid any impact of the effects of borderline in cases where a more open or closed NPD strategy is searched, only those observations, which are the same in four cases of all five, are considered powerful enough to indicate the existence of commonality. The validity of the overall observational judgement has been proven by writing a two- to threepage managerial summary of the new product strategy followed by each firm and by pointing out which performance terms were considered a strength and which a weakness accordingly. The case-specific summary has been mailed to the appropriate managing director. Each one has been later called by phone in order to ask whether they agreed or disagreed with the judgements of the researcher. No invalidating differences of opinion emerged. The realization of the indicated new product strategy and its corresponding implementation regarding the quality of new product processing has been proven also through a separate questionnaire. Its aim was to capture the characteristics of the latest five single new product efforts along the separate

225

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

phases of a conventional new product process. The questionnaire, basically, consists of a list of over 100 NPD statements characterizing the nature of each new product effort and was sent to the manager in charge of NPD a few weeks after the original in-depth interview. The manager was asked to evaluate the new product efforts by marking whether a certain statement was valid or not. Typical statements covered such aspects as whether a new product effort was based on a careful preplanning or not, whether there were a number of new ideas available, whether distributors participated in the effort, and what was the basis for starting a new product effort. Industrial SMEs representing conventional Finnish metal industry that had a minimum of 15 employees were selected as a preliminary target group from the list of members of the Federation of Finnish Metal, Engineering and Electrotechnical Industry. A final sample of five firms that had a turnover of e5 to e13.5 million and that were willing to participate in the research project, was formed, given the precondition that each firm had to be able to present as a minimum a clear history of three successive NPD rounds. All case firms are manufacturers of industrial machinery and equipment, which are sold to other industrial or business firms. The identified level of investment in single new product efforts varies between e42,000 and e605,000. The focus of the analysis is the NPD strategy and its implementation in the 1990s but the original ideas of the present business and corresponding products in all cases date back as far as the 1970s.

Empirical results Quality of new product strategy The identified parallels between the interview data and terms of performance are summarized in Table I. The data, which are available from the questionnaire have been used to validate what the interviewing data have indicated in terms of the NPD strategy. In this respect, a number of typical statements, against which the final characteristic of the prevailing NPD strategy of each firm is defined, are presented in Appendix 4.

Systematics As to the first strategic category, the systematic which indicates how firms plan their future new product efforts, case firms I, II and V state that they possess a capability to make strategic choices. The preset commonality criteria expressed in the last column of Table I ± only four observed cases of the total five are strong enough to indicate an existence of a commonality ± indicate nevertheless that the most common and visible aspect of a new product strategy is the lack of long-range planning which, on the other hand, can be considered as the very basic precondition of making relevant strategy choices. The lack of strategic planning applies more specifically to goal orientation and the clarity by which new product scope and the corresponding development path are defined. The interviewing data further indicate that instead of an explicit and predefined NPD strategy, emerging NPD needs tend to be defined as follows: . through NPD manager's own personal experience and knowledge and on ad hoc basis only as a reaction to the emerging changes in external circumstances; . along the commitment which is generated in top management discussions; and . by learning the consequences of earlier NPD efforts, which may justify the mobilization of any R&D resources available. Technology Given the second category, the technology, a high degree of external orientation and a more open NPD strategy is typically characterized by strong R&D orientation, search of useful knowledge from available external sources and of high technological product newness. In this respect, the identified lack of long-range planning may be due to the fact that firms, regardless of their level of systematics, believe that they have some reasonable search, development and application capabilities of some useful new technological knowledge. Research data however, clearly point out that high investment for the search of new technologies, in general, tends to be avoided. A strong technological orientation with a relatively high R&D spending and a highly open new product strategy can be observed in one case only, namely case number III. In all other cases, resorting more to their own

226

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

Table I Parallel matrix by new product strategy

resources, instead of searching for more collaboration with other useful knowledge holders, is more common, which indicates that a more closed new product strategy is typically followed. This may partly be due to the large application scope, which most firms state they possess because the utilization of the present scope does not necessarily require the establishment of a new product platform. The markets When it comes to the content of the established new product strategy, in terms of the ability to search, find, select and enter a market place, having characteristics which can be expected to drive a new product success, all case firms state that they can manage this facet of new product strategy without any specific difficulties and that good growth possibilities and sales potential exist. The inherent market fit to the firm's resources and competitive position in most cases is defined implicitly along the managers' own personal judgement. In the background of this state of affairs, the fact that in all cases the

key people involved have relatively long experience in their branch, has made it possible to specialize to a given product line and corresponding market niche. The research data indicate further that case firms typically stay close to their present customers, which is why the need of new business knowledge is not so urgent and compelling. This may also explain why the use of formal strategic planning tools is not considered a necessity either. In any case, a good market fit in relation to resources, which the firms in general can make available, is strongly emphasized by the interviewed managing directors. Resources The importance of new products is well acknowledged among the top managers who all believe NPD needs their support. However, case firms tend to avoid high NPD risks in spite of the fact that employees are in general encouraged to search new ideas and those managers are ready to bear the risks of NPD. The underlying data of single new product efforts clearly suggest that the search

227

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

of new ideas and any trials are mostly based on a short run and immediate customer response, which is available either through the distribution network or the prevailing customer intimacy.

Implementation of new product efforts Key observations on the NPD process The core of implementing a new product strategy is the process of how actual NPD is made and managed. In this respect, the parallels, which can be identified between the interview data and most important process factors the literature suggests, are summarized in Table II. Basis of the new product effort The terms of resource management concerns not only the identified top management support but they are met at the level of a given new product effort also. A minimum acceptable fit of new product efforts to the firm's resources can be observed in all cases. What is even more prominent is that firms are reluctant to resort to any formal and sophisticated NPD screening methods. The lack of proper screening procedures is related to the non-existent use of any dedicated and detailed market research. Firms simply trust

more their own in-house market knowledge or their distributors than any outside agency. The very few occasions where an outside market research agency has been involved are considered more or less disappointing, whatever the reason. Definition of the new product concept As for the definition of a new product concept, the terms of performance are met to a varying extent. The visible non-existence of relevant market research and corresponding detailed and credible market knowledge is especially reflected by the fact that new products are not positioned in a measurable way along the perceived mindset of customers. Partially, this seems to be due to the fact that in addition to more or less standard industrial products, case firms typically have had either in the past or have today tailormade engineering. In such a case detailed direct customer-specific knowledge, which can indicate which product attributes are of high value to the customer, becomes available. Such knowledge is most probably useful in terms of standard products, given that the knowledge is properly managed and shared among relevant people.

Table II Parallel matrix by new product process

228

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

New product management The management of new products varies a lot by case. Typical aspects are that the business targets of the new product efforts are not clearly stated. This applies to the financial aspects of the business targets too, given that the research data indicate that even the realized costs of new product efforts are not necessarily known or registered to such an extent as would be necessary to manage NPD effectively. From this point of view, the existence of two different sets of products, more or less tailor-made engineered and standard products, seems to have its drawbacks. It can be seen that if the division of labour between the two product categories is not made clear, the end result tends to be more engineering-driven than market-driven NPD because customers rarely participate in the actual development work other than in those cases where a new product is made on a tailor-made basis. In this respect, the core of the NPD becomes flexibility, by which necessary changes can be managed independently of the origin of the new product concept. More open or more closed strategy Based on the parallel matrixes and the managerial summaries, the identified NPD characteristics are summarized in Table III. The basis of the present business and application of NPD practices vary a lot case by case, leading to different performance results. The following key conclusions can be made: Case I The NPD is characterized by the fact that the product technology is known but it is applied in a combined way, which can offer unique product advantages. Applying core technology in the form of a product is demanding but in no specific way. One should also acknowledge that key managers have a good understanding of customer needs through their own experience. The firm is also in a favourable competitive position because leading industry players have not entered the market segments in question. It has been possible to follow a relatively closed NPD strategy, which has turned out to be highly successful mainly for two reasons. First, the case firm has been able to acknowledge and enter a new and emerging market. Second, a generic new product concept is in question and its application scope can be extended with minimal costs,

making NPD highly cost-effective. The above average performance becomes even more evident because the extension of the application scope can be based on a natural market growth in a market segment where no remarkable competitors exist. Case II The firm possesses a large cumulative knowledge base about matured technology, the application scope of which is gradually extended. Despite the maturity of the technology itself, its applications are such that some customer education is necessary in almost all cases to win a new customer. In this respect, the product customization takes place via the engineering department. New product efforts are mainly based on the search of new applications which existing cumulated experience makes possible. A relatively closed and reactive strategy is followed in the sense that the present technology scope directs the future new product efforts. Any considerable effort to enter new areas is avoided. Even if a profitable company is in question, an honest growth and performance level have been achieved in spite of the firm's ability to manage sophisticated and complicated product applications. Case III The case firm represents the most technologically oriented and proactive NPD, which relies on the ability to convert the latest technological knowledge into relevant product applications. A highly open NPD strategy is being followed which is based on an innovative core product. The product is protected by relevant industrial exclusive rights and forms the basis of the key business. The search of new knowledge is largely based on the cooperation with other useful knowledge holders. New markets and application possibilities are actively searched for in terms of the targeted growth. Sophisticated strategic planning tools are used in the search of new markets as well as in the formulation of the NPD strategy. Case IV The case firm is based on a derived business demand. A specific core product that serves the needs of one concentrated industrial branch only is in question. The relatively high dependence of the very few customers indicates why a closed NPD strategy has been the only one possible to apply. Given that a

229

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

Table III Characteristics of NPD quality by case Case No. I

II

III

IV

V

Basis of present business and key features of prevailing NPD practice An identified market gap and corresponding generic product concept with high substitution power is in question. Application scope is extendable by product accessories. NPD is made at minimum cost and in-house, mainly by combining existing knowledge Business is based on good cumulative knowledge of a mature technology and its application possibilities. Opening new market segments requires customer education. NPD is made mainly in-house, is largely sales oriented and has an engineering background The driving force behind the present business is a highly innovative core product having large substitution power. NPD is based on close cooperation and new knowledge is continuously searched A specific industry related product, based on a mutual business dependency, is historically in question. NPD is made strictly in-house and is dominated by the corresponding technical department The basis of the business is a unique product concept and the extension of the original product scope to other closely related areas. NPD is made mainly in-house and is largely based on a continuous generation of minor NPD efforts

functional specialization prevails throughout the organization, NPD is so far made strictly in-house and based on the firm's specialized knowledge. The firm has been able to develop technologically highly advanced products but with limited overall business performance. Case V The case firm has its origin in a unique product concept, which has made it possible to establish the business along the substitution power of the inherent innovation made. Due to the nature of the product innovation and the product technology utilized, the application scope is defined by the underlying product technology itself. Only strictly related application areas outside the core business can be targeted. Under these circumstances, a large variety of new product efforts with minimal change are ongoing. Even if cooperation exists and is sought, it takes place mainly in the area of practical engineering. Otherwise NPD is mainly made in-house, indicating that a more closed strategy as opposed to an open one, is

Characteristic of prevailing NPD strategy Cost-effective and closed strategy is followed. New market segments are opened via natural customer income and emergence of new demand. When measured by sales growth business performance is above average Relatively reactive and closed strategy is followed. New market segments are opened via continuous application search. When measure by sales growth business performance is modest

Technology-oriented, sophisticated and highly open proactive strategy is followed. Good market position in the industry is targeted. When measured by profitability overall business performance is average Closed and functionally oriented, new product strategy is followed. Larger business scope is targeted. When measured by economic competence performance is limited Conventional, relatively closed, but in a way a fishing new product strategy is followed. Continuous sales growth is targeted. When measured by sales growth business performance is fair

followed. New product features are expected to refresh the present sales and to generate business growth.

Review and discussion Firm behaviour in terms of NPD strategy Lack of resources and the nature of NPD are generally expected to make inter-firm cooperation crucial for many SMEs (e.g. Huang et al., 2001). Given that effective decision making depends on external information and linkages, firms who use more information are found to have achieved better results in the past (e.g. Li and Calantone, 1998; Freel, 1998). This applies to NPD also and becomes more important, when more innovative technological new products are in question (e.g. Dickson and Hadjimanolis, 1998). Innovative new products with a high technology content by definition require some useful new knowledge, the search of which

230

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

may not be possible without close cooperation (e.g. Cohen and Levinthal, 1990). A full-scale use of linkages and external information applies to one case firm only where cooperation is the key behind the strong technological orientation. Otherwise, such an orientation and search of corresponding new knowledge seems to be rare among SMEs. The case evidence indicates that more often a reactive and closed new product strategy is followed which corresponds with results found elsewhere (e.g. Soderquist and Chanaron, 1997; Freel, 1998). However, contradicting the views favouring open strategies, the case evidence expresses that, in one case, NPD can be highly successful regardless of the degree of cooperation. This supports the view that NPD is highly contingent (Link, 1987). In this respect, one may assume that strategic learning, which takes place locally, becomes a more important facet of new product performance and corresponding sales growth than an active search of new technological knowledge per se (c.f. Ledwith, 2000). The identified lack of strategic direction implies, on the other hand, that NPD is not made as effectively as it would be possible in the long run (c.f. Cooper and Kleinschmidt, 1996; Itami and Numagami, 1992; Pavitt, 1990; Maidique and Zirger, 1988; Gobeli and Brown, 1987). Even if the importance of NPD is acknowledged by top management in all case firms, there can be seen a certain tendency toward accepting NPD through necessity as a reaction to the prevailing market situation or NPD is based on a routine, instead of using innovation as a strategic tool. Avoiding higher risks may be justified on a case-by-case basis but it has its disadvantages also. There are reasons to expect that initiative conditions play a much larger role than it has been generally acknowledged regarding the new product performance in the long run. Even if it has been suggested that to be successful one should innovate for the present instead of the future (Drucker, 1985), the path-dependency of innovation implies that firms, who try to avoid higher risks, face a threat of being locked in a pattern of following reactive NPD only (e.g. Lindman, 2000). The locking effects of the initiative conditions in which the original market entry has been made will largely define the future growth rate of the firm even in the long run (c.f. Pavitt, 1990). Some tentative evidence of this idea can be seen in the fact that,

in case firms, it has taken roughly some 20 years to reach the present level of turnover, given that a more closed NPD strategy has been followed. The firm behaviour in terms of NPD implementation As to the implementation of new product efforts, the underlying concept of marketing orientation, having reasonable market knowledge and marketing skills, identifying the real customer needs or making any market research, has been generally considered one of the key success factore (e.g. Brown and Ennew, 1995; Griffin and Hauser, 1996; Calantone et al., 1993). In this respect, case firms are representatives of the most poorly managed facet of NPD. Resorting to their own in-house market knowledge solely or to the feedback received from an existing distribution network ± a typical means to assure the fit of a new product concept in all cases ± follows in line of its reactive new product approach. This implies that in the long run, SMEs are apt to become prisoners of their own strategy due to the cumulative impact of NPD on the firm's competence base (e.g. Maidique and Zirger, 1988). In addition to losing the learning efforts of a more innovative NPD, there is the risk of developing yesterday's products. This is simply due to the natural fact that customers' views and any corresponding new product opportunities are based on the experience they have of the present products (Teece et al., 1997). Given the acknowledged inability or reluctance to make any exact product positioning, which was found in all cases, one may expect some considerable countereffects, which are met later on during the launch phase. A good and well-defined product positioning can indicate which attribute set is of greatest importance to the customer ± thus having a considerable effect on the new product design. It also offers a clear basis on how to differentiate one's new product offer from the competitive ones (e.g. Urban and Hauser, 1996). It should be noticed as well that new product positioning when properly made can be a valuable tool in terms of generating any sales argumentation and advertising the final new product. The fact that NPD strategy tends to be implemented with minimum formalism can be seen in the background of when flexibility becomes the factor through which case firms can manage the final acceptable end result of

231

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

their new product efforts. This agrees with the earlier research indicating that flexibility in general has a relatively prominent impact on new product success (e.g. Ledwith, 2000; Davidson et al., 1999). The perceived flexibility, extended experience and corresponding cumulative customer knowledge which the case firms possess in their branch can be seen as the driving force of the new product fit which is reasonably achieved between the market and their own resources. Like flexibility in general, the market fit in relation to resources has been considered an important success factor in the earlier literature, especially in regard to any inherent synergy effects (e.g. Cooper, 1987). Final notes on the identified quality of NPD One has to point out the diversity of the context and initiative conditions where case firms have their origin. All firms have kept going continuously through a number of casespecific events, which have in many ways restricted the use of given resources. In this respect, the NPD history of case firms closely resembles the concept of muddling through (e.g. Noda and Bower, 1996). Given the increasing trend of intra-firm cooperation, a more open strategy and, accordingly, increasing cooperation is called for if the benefits of the network approach are agreed upon. In this respect, the findings of the study indicate that SMEs can continuously achieve reasonable new product performance by following a reactive and more closed NPD strategy and by largely relying on their own resources only.

References and further reading Bengtsson, M. and Kock, S. (1999), ``Cooperation and competition in relationships between competitors in business networks``, Journal of Business & Industrial Marketing, Vol. 14 No. 3, pp. 178-93. Brown, A.D. and Ennew, C.T. (1995), ``Market research and the politics of new product development``, Journal of Marketing Management, Vol. 11 No. 4, pp. 339-53. Calantone, R.J., DiBenedetto, A. and Divine, R. (1993), ``Organizational, technical and marketing antecedents for successful new product development``, R&D Management, Vol. 23 No. 4, pp. 337-51. Christensen, C. (1997), ``Patterns in the evolution of product competition``, European Management Journal, Vol. 15 No. 2, pp. 117-27.

Cohen, W.M. and Levinthal, D.A. (1990), ``Absorptive capacity: a new perspective on learning and innovation``, Administrative Science Quarterly, Vol. 35, pp. 128-52. Cooper, R.G. (1987), ``The performance impact of product innovation strategies``, European Journal of Marketing, Vol. 18 No. 5, pp. 5-54. Cooper, R.G. (1998), ``Benchmarking new product performance: results of the best practices study``, European Management Journal, Vol. 16 No. 1, pp. 1-17. Cooper, R.G. and Kleinschmidt, E.J. (1996), ``Winning business in product development: critical success factors``, Research Technology Management, Vol. 39 No. 4, pp. 18-29. Davidson, J.M., Clamen, A. and Karol, R.A. (1999), ``Learning from the best new product developers``, Research Technology Management, Vol. 42 No. 4, pp. 12-19. Dickson, K.E. and Hadjimanolis, A. (1998), ``Innovation and networking amongst small manufacturing firms in Cyprus``, International Journal of Entrepreneurial Behaviour & Research, Vol. 4 No. 1, pp. 5-17. Drucker, P.F. (1985), Innovation and Entrepreneurship, Heinemann, London. Freel, M.S. (1998), ``Barriers to product innovation In small manufacturing firms``, paper presented to 21st National Small Firms Conference, Durham, November, pp. 1054-72. Gobeli, D.H. and Brown, D.J. (1987), ``Analyzing product innovations``, Research Management, Vol. 30 No. 4, pp. 25-31. Griffin, A. (1997), Drivers of NPD Success: The 1997 PDMA Report, PDMA, Chicago, IL. Griffin, A. and Hauser, J.R. (1996), ``Integrating R&D and marketing: a review and analysis of the literature``, Journal of Product Innovation Management, Vol. 13, pp. 191-215. Huang, X., Soutar, G.N. and Brown, A. (2001), ``Resource adequacy in new product development: a discriminant analysis``, European Journal of Innovation Management, Vol. 4 No. 1, pp. 53-9. Itami, H. and Numagami, T. (1992), ``Dynamic interaction between strategy and technology``, Strategic Management Journal, Vol. 13, pp. 119-35. Ledwith, A. (2000), ``Management of new product development in small electronics firms``, Journal of European Industrial Training, Vol. 24 Nos. 2-4, pp. 137-48. Li, T. and Calantone, R.J. (1998), ``The impact of market knowledge competence on new product advantage: conceptualization and empirical examination``, Journal of Marketing, Vol. 62, October, pp. 13-29. Lindman, M. (2000), ``New product uniqueness in the context of industrial product development``, Journal of Marketing Management, Vol. 16 Nos. 1-3, pp. 247-71. Link, P.L. (1987), ``Keys to new product success and failure``, Industrial Marketing Management, Vol. 16, pp. 109-18. Maidique, M.A. and Zirger, B.J. (1988), ``The new product learning cycle``, in GroÈnhaug, K. and Kaufmann, G. (Eds), Innovation: A Cross-Disciplinary Perspective, Norwegian University Press, Oslo.

232

Open or closed strategy in developing new products? A case study

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Martti Tapio Lindman

Noda, T. and Bower, J.L. (1996), ``Strategy making as iterated processes of resource allocation``, Strategic Management Journal, Vol. 17, pp. 159-92. NystroÈm, H. (1990), Technological and Market Innovation, John Wiley & Sons, New York, NY. Pavitt, K. (1990), ``What we know about the strategic management of technology``, California Management Review, Vol. 32 No. 3, pp.17-26. Pavitt, K. (1991), ``Key characteristics of the large innovating firm``, British Journal of Management, Vol. 2, pp. 41-50. Rothwell, R. (1983), ``Innovation and firm size: a case for dynamic complementary, or is small really beautiful``, Journal of General Management, Vol. 8 No. 3, pp. 5-25. Rothwell, R. (1992), ``Successful industrial innovation: critical factors for the 1990s``, R&D Management, Vol. 22 No. 3, pp. 221-39. Soderquist, K. and Chanaron J.J. (1997), ``Managing innovation in French small and medium-sized

enterprises: an empirical study``, Benchmarking for Quality Management & Technology, Vol. 4, pp. 259-72. Teece, D.J., Pisano, G. and Shuen, A. (1997), ``Dynamic capabilities and strategic management``, Strategic Management Journal, Vol. 18 No. 7, pp. 509-33. Urban, G.L. and Hauser, J.R. (1996), Design and Marketing of New Products, Prentice-Hall, Englewood-Cliffs, NJ. Yin, R.K. (1989), Case Study Research, Design and Methods, Sage Publications, Newbury Park, CA. Zineldin, M. and Jonsson, P. (2000), ``An examination of the main factors affecting trust/commitment in supplier-dealer relationships: an empirical study of the Swedish wood industry``, The TQM Magazine, Vol. 12 No. 4, pp. 245-65.

(The appendices follow overleaf.)

233

Open or closed strategy in developing new products? A case study

Martti Tapio Lindman

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Appendix 1 Table AI New product performance factors by selected literature The author(s)

Year of publication

Cooper, R.G.

(1979)

Cooper, R.G.

(1985)

Cooper, R.G.

(1986)

Link, P.L.

(1987)

Cooper. R.G. and Kleinschmidt, E.J. Cooper, R.G. and Kleinschmidt, E.J. Kleinschmidt, E.J. and Cooper, R.G. Rothwell, R.

(1986) (1987) (1988) (1992)

Cooper, R.G. and Kleinschmidt, E.J.

(1993)

Calantone, R.J., Di Benedetto, A.C. and Divine, R. Calantone, R., Vickery, S. and DroÈge, C. Cooper, R.G. and Kleinschmidt, E.J.

(1993)

Cooper, R.G. and Kleinschmidt, E.J.

(1995)

(1995)

(1996)

Ayers, D., DahlstroÈm, (1997) R. and Skinner, S. Poolton, J. and (1998) Barclay, I. Cooper, R.G. (1998) Davidson, J.M., Clamen, A. and Robin, K. Cooper, R.G.

Summarizing articles Johne, A.J. and Snelson, P. Craig, A. and Hart, S. Montoya-Weiss, M. and Calantone, R. Griffin, A. and Hauser, J.R.

(1999)

(1999)

(1988) (1992) (1994) (1996)

Indicated performance factors

Research context

Product uniqueness and superiority. Excellent market knowledge. Marketing skills/effective launch. Technical and production synergy Highly technological, innovative, technically complex and proactive products. Strong market orientation. Search for potential and promising markets, which have high growth expectations and fewer competitors Identification of customer needs. Highly differentiating new product benefits. Strong R&D input. Synergy between the R&D and research projects Implementation of the launch. Synergy with the existing business. Complete market knowledge. Product/market attraction. Newness of the product. Quality of the product Management of new product development stages, especially the preliminary market assessment and formal launching Product benefits. Careful management of pre-development stages. Careful definition of the new product effort International market orientation from the very beginning

Single product

Good internal and external communications. Innovation throughout the organization. Good management and management style. Product champion. Good planning and control. Efficient new product development work. Understanding customer needs Quality of new product development process. Early and sharp definition of the new product. Strong team leadership and multifunctionality. Customer-based new product idea. Quality of the launch Flexible, non-mechanical organization. Marketing skills and technical activities

Innovation orientation of top management. Participation of the production in the development (assuring the product flexibility). Decentralization of new design and development New product uniqueness, innovative products. Multifunctional teams. Well-done homework. Quality of marketing activities. Product definition High quality development process. Clearly defined new product strategy. Sufficient human and financial resources. Sufficient investments in R&D. Qualified new product teams. Top management commitment. Innovative corporate culture. Cross-functional teamwork. Controllable results Integration of R&D and marketing. Perceived management efficiency Top management support. Long run new product plan. Flexibility and ability to change. Risk taking ability of top management. Entrepreneurial culture Qualified systematic new product practice. Clearly stated new product goals and strategy. Sufficient support and investment in relation to new product targets Clarity and clear practices. Ownership and commitment. Leadership and commitment among the top management. Integration. Flexibility and willingness to change Well-done up front homework. Listening to the customer's voice. Superior, differentiated products. Sharp and early definition of new product concept. Launch planning from the very beginning. Stage gate development. Multiskills, cross-functional teams. International orientation

The whole new product program (the last five years) The whole new product program (the last five years) Successful/unsuccessful products Management of new product development Launched product Single product New product development in general

New product activities and their performance at a product level (chemical industry only) Organization structure and its impact Eight key new product activities of furniture manufacturers Successful/unsuccessful products (chemical industry only) New product development by business units

Organizational preconditions (computers only) Innovation in general New product development by business units Management of new product process Single product

``Success factors in product innovation: a selective of review of the literature'', The Journal of Product Innovation Management ``Where to now in the new product development research'', European Journal of Marketing ``Determinants of new product performance: a review and meta-analysis'', Journal of Product Innovation Management ``Integrating R&D and marketing: a review and analysis of the literature'', Journal of Product Innovation Management

234

Open or closed strategy in developing new products? A case study

Martti Tapio Lindman

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

Appendix 2 Table AII Evaluation basis of NPD New product performance factor

Terms of performance

Systematics Visionarism Long-range planning Goal orientation Clarity Toughness Communication

Capability to make strategic choices There is a long-range plan for new products Objectives and roles of new products have been defined in terms of business goals Business scope and corresponding new product development path have been defined in the long run Ability to invest in new products persistently Meaning and importance of new products have been made clear to all people in the organization

Technology Role of R&D in business strategy Character of the products developed Degree of innovation Generity of the product Matchlessness

Firm is strongly R&D oriented and searches actively for new technological knowledge Ability to search, develop and apply new knowledge Products have technological newness Products have a large application scope Product uniqueness is searched for

The markets Market orientation Nature of target markets Market growth Market size Competition International orientation Market knowledge Timing of market entry Customer expertise

Ability to search and find potential markets Fit of markets to the firm resources Markets have a considerable growth potential Markets have a considerable sales potential Ability to compete in the market Products are planned from the very beginning for international markets Length of market experience in the branch Ability to pioneer in the market Knowledge of customer needs and user conditions

Resources Commitment Culture Organizational business logic Resource inputs

Top management supports new product innovation Search for new ideas and trials is encouraged Firm has customer-based management system Top management is ready to take risks and invest as much as necessary to achieve the business goals

Basis of new product efforts Overall impact of the effort of the firm Screening method Preliminary market assessment Preliminary technical assessment Preliminary business assessment

Fit of NPD efforts to the firm resources Selection of new product efforts is based on clear criteria developed and agreed on in advance New product efforts are based on detailed and credible market knowledge Costs, risks and timing of new product efforts are mapped as to a new product and its production Objectives of new product efforts have been set out and financial end results estimated

Definition of the product concept New product competitiveness Market requirements Product definition Product positioning

New product concept contains clear and visible product attributes which increase competitiveness Target markets and their characters are defined Product specifications have been defined in full New product position along the customer mindset and preferences has been defined

New product management Management style Goal setting Development pattern Organizing of new product efforts Flexibility

There is a demand for a highly qualitative process New product efforts must fulfil clear business prerequisites NPD is managed by using a systematic development protocol and procedures NPD is managed by using systematic functional development teams It is possible to make changes during the process

235

Open or closed strategy in developing new products? A case study

Martti Tapio Lindman

Appendix 3. List of persons interviewed by professional status Case I Managing director NPD manager Production manager Case II Managing director NPD manager Production manager Marketing manager Case III Managing director NPD manager Production manager Sales manager

European Journal of Innovation Management Volume 5 . Number 4 . 2002 . 224±236

production accepts too easily that one can have a basket which is 20mm lower or higher or with handles here and there or would you please promise me a basket which is 30mm deeper even if standard equipment always is in question. We have largely made the product development in such a way that it has taken place through the sold machinery projects (production manager). We should be able to identify and study better what it is really needed in the market place next year or the second or even the third year ahead (managing director). We have not basically had any cooperation. Only in one technical area we have a couple of subsuppliers with which we do development and testing. But as to other technical matters there is no cooperation (NPD manager).

Case III We have not had any special need to resort to external resources in NPD. We do utilize language specialists, to have some documents translated. Perhaps there may be something else, too? However, I think mostly these are the things, may be there is nothing else (NPD manager). Well, we have, even today, we used industrial designers with this outlook . . . They are external people, it is bought outside the house. At the moment we buy some amount of external engineering planning but, mainly, the development work is made by ourselves (production manager).

Case IV Managing director NPD manager Production manager Marketing manager Case V Managing director NPD manager Production manager Service manager Engineering planners (two persons) Marketing manager

Case IV Going back to the question of which people are involved and think about the strategy, it is mainly myself and the management team . . . It is natural to use subsuppliers and also it is natural to utilize partners in the development work . . . . It is basically the only opportunity (managing director). Let us say, we search development trends in terms of what we should do, but then, what is actually made, here the external quarters are very strongly involved with (NPD manager).

Number of persons interviewed 21, in total

Appendix 4. Typical statements indicating the degree of openness/ closedness of the prevailing NPD strategy Case I So far we have not used any external designers . . . . We have had some connections with the technical high school. We have however come to the conclusion that in a certain way they are too far away from the application phase. They are thinking too much in the long run . . . . When you were talking about the NPD costs in the very beginning, it is so that we have quite a small number of people doing this (NPD manager). We have done in principle everything ourselves. We have not really been able to enter into cooperation with the high school in such a way that it would have been fruitful. They consider things differently. We have done ourselves the whole new product development (managing director).

Case II Well, let us say that because standard products are in question, the sales, this representative of

236

Case V When we have been able to describe the problem we then collect a team of professors, ``village blacksmiths'', drawers, designers, engineers, in total five to six persons. It is that work once we have been able to solve the problem ourselves at first (NPD manager). The absolutely largest part of our NPD is product improvements (marketing manager). We do not use external resources to any notable extent. Product development as such utilizes various designers and specialists when they have some problem. But for production, we do not use (production manager). We know approximately how much a NPD project will cost. Whether it will succeed or not it does not matter . . . In fact our projects are not so large that they would present a risk to the firm's health or actions (managing director).