The New Product Development Process as a Communication Web, Part II: Analysis of Spanish Firms
Universidad del Pais Vasco, Spain
University of Santiago de Compostela, Spain
University of Santiago de Compostela, Spain
Universidad del Pais Vasco, Spain
In the previous chapter (Part I), we proposed a model relating the composition and external communication activities of NPD teams to the performance of NPD programmes. In this chapter (Part II), through the use of structural equations analysis, we compare the model to a sample of 136 managers from different functional areas at 121 innovative Spanish firms. The results indicate that the impact of explanatory variables on new product programme performance differs according to the measure of performance considered. The cross-functional nature of NPD teams, the presence of product champions in NPD teams and the gathering of information by all NPD team members were all shown to positively influence new product performance. Firms should be aware of the importance of the aforementioned variables.
Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
In Part I of this chapter we suggested considering New Product Development (NPD) as a communication web, contributing towards the innovation activities of companies and providing them with a sustainable competitive advantage. The external communication activities and cross-functional nature of NPD teams positively influences new product programme performance and provides companies with attributes that competitors find difficult to emulate.
The aim of this chapter is to compare the proposed model and examine whether or not new product programme performance is influenced by: (1) the cross-functional nature ofNPD teams; (2) the presence of product champions in the NPD process; (3) the presence of gatekeepers in the NPD process; and (4) NPD lead users.
We tested the aforementioned model using a sample of 136 managers from different functional areas at innovative Spanish firms. The results obtained from a structural equations analysis indicated that the impact of explanatory variables on new product programme performance differs according to the measure of performance considered.
Our study contributes towards existing NPD literature as, unlike other studies on the communication web approach, it takes a development programme of three years as its unit of analysis and examines explanatory variables whose effects on performance have not yet been studied together, in firms that belong to different sectors and can provide data from cross-functional sources.
The chapter is structured as follows: firstly, we recapitulated the hypotheses proposed in Part I. Secondly, we described the method used, tested the model and commented on the main findings. We subsequently identified the implications ofthe findings for NPD managers and, finally, discussed the chapter’s limitations, as well as possible future lines of research.
Based on previous evidence (Clark & Fujimoto, 1990, 1991; Katz & Tushman, 1981; Markham & Griffin, 1998; Von Hippel, 1986), we propose that new product programme performance will be influenced by the cross-functional nature ofNPD teams and their external communication activities. Specifically, we hypothesise the following:
H1: The cross-functional nature ofNPD teams, which is measured by the number of departments participating in the NPD process, will positively influence new productprogramme performance.
H2a:The presence ofinformation gatekeepers will positively influence new productprogramme performance.
H2b: The impact on new product programme performance will be greater with the presence of information gatekeepers than when all members of the NPD project team are in charge of gathering external information. H3: The presence of innovation champions will positively influence new productprogramme performance.
H4: The participation of lead users in NPD will positively influence new productprogramme performance.
The sample was selected from the innovative firms database of the Centre for Technological Development of Industries (CDTI): an organisation that promotes the innovation and technological development of Spanish firms. In order to form part of the sample firms had to meet two requirements: belong to one of the industries shown in Table 11 and have two or more people dedicated to R&D tasks. These conditions were met by 600 of the firms.
Two questionnaires were posted to each ofthe 600 firms, one addressed to the R&D manager and the other to his/her co-worker in the commercial or marketing department. A total of 121 firms (20.2 per cent of those contacted) responded to the questionnaires. Although the letters were addressed to the R&D and marketing managers, in some cases the responses came from managers in other areas2. A total of 136 questionnaires were received from 121 companies.
All the measurement scales are subjective and reflect the perceptions ofmanagers from different areas (above all from R&D and marketing departments) on new product programme performance in the last three years, the average number of departments participating, the presence of information gatekeepers and product champions and, finally, the consideration of lead users.
To measure new product programme performance we used the scale designed by Cooper (1984), consisting of six items or indicators corresponding to three dimensions: the overall programme performance (three items), the impact of
Table 1. Industries in the sample
theprogramme on thefirm (two items) and the new products overall success rating (one item). The measurement scales for the variables (presence of gatekeepers, presence ofproduct champions, consideration of lead users, physicalproximity and cross-functionality) were developed for the study on the basis ofthe definitions of previous studies.
The measurement scale used for the presence of gatekeepers was based on the evidence presented in previous research papers regarding the benefits of using a limited number of key people within the development project in order to obtain external information, as opposed to all team members being equally responsible for keeping up with changes in the environment (Allen and Cohen, 1969; Allen, 1970). For this reason, rather opposing suggestions were made (seeAppendix):
“In my firm there are one or more people who gather and interpret external information on changes in the environment and make it more accessible for us (V9).”
“All members of the NPD project teams are expected to keep up with changes in the environment” (V10).
The one-item measure used for the presence of product champions was based on the principal characteristics ofthis figure as highlighted in previous research papers. Therefore, the managers of different departments were asked to express how much they agreed or disagreed with the following statement:
“In each ofour project teams there is at least one person who actively and enthusiastically drives forward each NPD stage, takes risks and does not give up when faced with obstacles” (V8).
Consideration of lead users in NPD is summarised by the variable V11: “To seek out market opportunities the firm pays special attention to “pioneering” consumers (i. e. those who identify general market needs, but identify them months or years before the majority of the market). For this measure we took into account the definition by von Hippel (1986), but opted to replace the term “lead users” with that of “pioneering consumers,” in case managers were unfamiliar with the former.
The gatekeeper, product champion and lead user scales were reviewed by 14 managers, who did not observe any inconsistencies in the measures proposed. These variables were measured by Likert type scales of seven points (1= totally disagree; 7= totally agree)
Finally, the cross-functional nature of NPD teams was measured using managerial estimates of roughly how many departments would participate in the NPD teams in the following three years (V 7).
The industry with most cases in the sample is that ofthe chemical sector (50 cases), followed by the electronics sector (26 cases), the machinery construction sector (24 cases) and finally the plastic manufacturing sector (21 cases).
The means and standard deviations of the indicators of the study enable us to verify that, in general, managers strongly agree that: (1) the NPD programme greatly influences company sales and profits; (2) the objectives of the programme have been achieved; (3) costs have been covered by generated profits; and (4) the programme has been a success with regards to overall profitability (see Appendix).
In mean values, approximately one third of the firms’ sales come from products launched onto the market in the last three years and more than half (59 per cent) of the new products have been a commercial success in the last three years.
We removed four items from the new product performance measurement scale (the only one with various indicators). The items’ corrected item-total correlation was below the recommended
0. 3. The alpha value rose substantially when only V1 and V2 were considered (alpha = 0.82). The percentage of variance extracted on carrying out a principal components factor analysis with these two indicators was very high (84.72%).
Therefore, from the NPD programme performance scale initially proposed (i. e. one with six indicators), a two-indicator scale was developed: (1) from an overall profitability standpoint, our new product development programme has been successful; and (2) the overall performance of our new product programme has met our objectives. The proportion of new, commercially successful products in the last three years and the proportion of sales as a result of new products were only slightly related to the other indicators on the scale.
In order to somehow capture the three dimensions of the performance of new products scale proposed by Cooper (1984), we considered three different measures. Performance 1 corresponds to the overall programme performance and has the two indicators that passed the reliability analysis; Performance 2 measures the impact of the programme on the firm; and Performance 3 is an overall success rating of new products.
Convergent validity of the only scale with more than one indicator was examined using exploratory factor analysis. The items substantially load onto their respective factors. The loads ofthe two indicators ofthe first measure ofnew product performance (Performance 1: overall programme performance) are equal to 0.92, which indicates convergent validity.
To test the hypotheses we proposed a structural model that considered the influence of four independent variables: cross-functional nature of the team, presence of product champions, presence of gatekeepers and consideration of lead users on new product performance.
Table 2. Structural equation model results: parameter estimates (Performance 1)
Firstly, an estimate of the model (using the AMOS 4.0 programme) had to be made, considering the first measure of new product performance (Performance 1), and as a measure ofthe presence of gatekeepers a single indicator (V9). The crossfunctional nature of the NPD team, measured by the number of departments that usually participate in the NPD process, was excluded from the analysis after it was found that it did not affect performance. The gatekeeper variable was not significantly related to performance, while the gathering of external information by all members of the NPD project teams (V10) did positively influence performance p < 0.1 (see Table 2). The relation of the other two independent variables (consideration of lead users and presence of product champions) with new product performance was significant for p < 0.01.
The presence of champions positively and substantially influenced performance (b = 0.52), while consideration of lead users when developing a new product (contrary to what we presumed) is negatively related to performance (b = -0.22). This is paradoxical and contradicts evidence presented in former studies (Von Hippel, 1986; Herstatt and Von Hippel, 1997).
In view of the above results, and considering the first measure of new product performance (Performance 1) (see Table 2 and Figure 1), we can accept H3 (the presence of champions positively influences performance), but cannot accept the three remaining hypotheses, as neither the cross-functional nature of the team (H1), the presence of information gatekeepers (H2a), nor the consideration of lead users (H4) favour the programme’s overall performance. We cannot accept H2b either because the presence of information gatekeepers does not benefit performance more than the collection of external information by all team members. In fact, the opposite appears to be true (i. e., the gathering of information by multiple parties favours the performance of the NPD proj ect) whereas the presence of information gatekeepers has no influence on performance.
Figure 1. Structural model for the first measure ofperformance (Performance 1)
Table 3. Structural equation model results: Parameter estimates (Performance 2)
Figure 2. Structural model for the second measure ofperformance (Performance 2)
When the measure of performance in question is the proportion of total sales thanks to new products (Performance 2), only two out of the initial four variables are statistically and significantly connected to new product performance (i. e. p < 0.1) (see Table 3 and Figure 2). Given that R2 is low, the input of development teams, as well as of the gathering of information from the external environment, are fairly limited, but still significant.
When considering “the proportion of commercially successful new products in the last three years as a measure of performance (Performance 3: V6), two variables are distinctly connected to new product performance (V8 and V11), but the model is not identified (as there is not enough scope in order to estimate it).
With the aim of solving this problem, we recommend a multiple regression with Performance 3 (V6) as a dependent variable and V8 (champions) and V11 (lead users) as independent variables.
Corrected R squared
a Independent variables (Constant term), V8 (champions); V11 (lead users).
Table 5. Estimatesa
a Dependent variable: Performance 3.
The analysis carried out using the SPSS statistical programme is shown in Table 4 and Table 5. Although R2 is not very high, the coefficients of the independent variables are significant and show: on the one hand, that the presence of product champions positively influences the proportion of new products that have achieved commercial success in the last three years; and, on the other hand, that consideration of the needs of lead users negatively affects new product performance.
The cross-functional nature of NPD teams (measured by estimating the average number of departments participating in the NPD process) positively influences one of the measures of new product performance (i. e. the total percentage of sales represented by new products) (Performance 2), but does not influence the othertwo measures (overall programme performance and new products overall success rating). The frequency distribution of this variable allows us to confirm that only three managers use single-function teams within their firms, eight managers make use of the participation of two departments, and 97 (out of a total of 130 responses) use cross-functional teams from three to five different departments (the number of managers that use cross-functional teams from five departments is considerably low). Given that virtually all contacted firms use cross-functional teams for NPD, we cannot verify the advantages of this option over that of single-function teams. The variable is therefore not proven to be significant when it comes to explaining performance.
As a result, it can be concluded that innovative Spanish firms have faith that teams formed by people from different functional areas can carry out the development of new products.
A direct and positive link between managers’ views on the presence of product champions in NPD teams and the performance ofthe programme has also been found to exist, both when overall programme performance was considered (b=0.5 2) and when success ofnew products was considered (b=0.30) (see Table 6).
Table 6. Summary of the main results
These results are consistent with previous studies (Markham, 1998; Markham and Aiman - Smith, 2001; Markham and Griffin, 1998) and highlight the importance of product champions. For some authors, cross-functional teams and the formalisation of NPD processes do not eliminate
the need for product champions, “those passionate individuals who believe in the stated innovation strategy ofthe organisation and recognise the potential in an idea or opportunity” (Markham and Aiman-Smith, 2001, p.47).
The gathering of information from the environment by all the members of NPD teams was also shown to positively influence performance, especially overall programme performance and sales of new products (b = 0.17 in both cases). These results are supported by previous studies (Ancona, 1990; Ancona and Caldwell, 1997).
However, the presence of information gatekeepers was not significantly related to new product performance, contrary to what previous studies have found (Allen, 1970; Katz and Allen, 1981; Lievens and Moenaert, 2000). This could be due to the measure used or to the characteristics of the sample. However, we believe that, in addition to the aforementioned reasons, there are also other reasons.
Allen (1966) verified that, despite the apparent benefits associated with consulting colleagues, in the projects studied, more members resorted to obtaining ideas from external sources than from within their own organisations by simply consulting a member of technical staff. The performance achieved after acquiring information from outside was, however, rather poor. It is ironic that a high number of employees used sources of information that ultimately led to a poorer performance. Perhaps they did so as they felt embarrassed about admitting to one of their colleagues that they needed help in solving a problem.
It may also be the case that gatekeepers are not as equally valuable in some situations as they are in others. For Katz and Tushman (1981) the key can be found in the distinction between research projects or projects with a universal orientation, and development projects or projects with a local orientation.
When technological activities have a local perspective, individuals that take part in them share the same language, and communication is quick and easy. However, the acquisition and interpretation of information from external sources that use another language, makes communication more difficult. Projects with a local orientation require gatekeepers to establish a link with external areas, meaning contact with the outside environment is indirect.
Projects with a universal orientation, on the other hand, use the same language that is used externally. As there are no communication barriers there is also no need for gatekeepers. Communication with the outside occurs by all members of the group directly. Overall, research projects without gatekeepers and development projects with gatekeepers were associated with higher performance.
In their recent study, Whelan et al. (2010) questioned the individual role of gatekeepers. These researchers recommend that the gatekeeper concept be re-examined in view of the latest advances in Internet technology, which have significantly changed the way people acquire and share information. Although we may think that thanks to the Internet the circulation of and access to information on technological advances is within reach for all members of new product development teams, Whelan et al. (2010) found that, even though searching for information is a great deal easier due to the Internet, the verification, translation and internalization of such information requires skills held only by a select few. Whelan et al.’s case-study on a medical devices company demonstrates how the role ofgatekeeper can divide a workforce into two: the external communication stars, who are in charge of searching the Internet, informing the group of the latest technological developments and verifying the reliability of the information before discussing it with the internal communication stars, in turn, are in charge of finding a possible use for such information within the group and translating it into comprehensible terms for those that intend to use it. This kind of work presents new and interesting opportunities in relation to future research.
Finally, our findings indicate that Spanish firms’ employment of lead users in NPD negatively influences two of the three measures of performance considered, mainly the overall programme performance and the new product success. These results do not correspond to those of previous studies, which draws attention to the incorporation ofinformation from consumers into the project team (Hong et al., 2004; Luthje and Herstatt, 2004; Morrison et al, 2000; Von Hippel, 1986; Von Hippel et al, 2000), but they are consistent with the research of Song et al. (2006), who found that the use of lead user networks is negatively associated with the acquisition of new NPD knowledge.
The lead users method is proposed as a solution for those firms that wish to carry out radical improvements but that, single-handedly, are only able to develop line extensions and incremental innovations. Von Hippel et al. (2000) state that the lead user method is used in eight out of the 55 3M firm divisions and those cross-functional teams that take lead users’ opinions into account are very satisfied with the result. However, these researchers also posit that the use of this method does not guarantee success. A lack of management support or a team’s inability to carry out the process can cause project failure.
We believe that this negative impact of the consideration of lead users on NPD could be due to several reasons:
• Firms may not be following an appropriate method/process for studying company needs and identifying users. The method proposed by von Hippel (1986), developed in later studies (Luthje and Herstatt, 2004; von Hippel et al., 2000) discusses the need for development projects with lead users to go through four phases: (1) initiation and establishment of bases; (2) determination of tendencies; (3) identification of lead users and (4) development of product concept. Enkel et al. (2005) also put forward four phases to involve the customer in the NPD process. These researchers argue that companies need to understand how and when to use a lead-user approach in order to reduce the risks involved in radical innovations. In this respect, the Internet provides ways to incorporate potential customers into the NPD process. Active participation by NPD managers in virtual communities of bloggers and readers who are interested in certain types of products could also provide very valuable information in relation to the main marketing-mix variables (product, price, channel and promotion) and assist in the identification of lead users and early adopters (Droge et al., 2010). Additionally, the use of virtual stock markets or information markets could help to identify those lead users with a greater ability to forecast the market success of new products (Spann et al., 2009).
• One of the method’s theories may not be accurate. It may assume that the perceptions and preferences of lead users are similar to those held by non-lead users once the market for a particular product has developed. However, current users could decide that they do not like the product concept that lead users have helped to develop. In this case, one of two possibilities are likely to arise (Urban and von Hippel, 1998):
a. The new concept will not be appreciated immediately, but will be in the future when the needs of non-lead users have evolved and come to resemble those presented earlier by lead users.
b. The concept will never be appreciated by non-lead users.
• The involvement of lead users in the generation of ideas for new products can be very useful if a company is seeking radical innovations. However, if a company requires incremental changes, the use of ordinary users through the guided user approach (where information on the technology behind the new product is given) may be more effective (Magnusson, 2009).
We believe it is essential for future studies to analyse the link between lead users and performance. Should firms study lead users and incorporate their needs and ideas into product improvements or new products? Do firms have a long-term outlook? Are they willing to wait for the market to appreciate the products designed on the basis of lead user contributions? Is it easy to identify lead users? Are firms capable of using the lead user method? These and other questions need to be answered, particularly in relation to Spanish firms.
This chapter gathers the results of a study on a sample of innovative Spanish companies. It forms part of a line of research that regards NPD as a communication web. It posits that new product programme performance will be greater in firms that use cross-functional teams, rely on the presence of product champions and information gatekeepers and take the opinions of more advanced users into account.
The results, obtained from a sample of 136 managers from different functional areas (principally R&D, marketing and manufacturing) partly support the hypotheses put forward, and allow us to conclude that:
1. The three dimensions proposed in the measure of new product programme performance are perceived by managers as three distinct measures: (1) the overall programme performance; (2) the impact of the programme on the firm and (3) an index of the success of the new products.
2. The use of cross-functional teams in NPD is a general practice in innovative Spanish firms and positively influences the impact of the programme on the firm.
3. The presence of champions is important, both in achieving good overall programme performance and in obtaining a high successful product rate.
4. The collection of outside information by all members of cross-functional teams helps to improve overall programme performance and its impact on a firm’s results.
5. The presence of information gatekeepers is not associated with any ofthe three measures of performance.
6. The consideration of lead users in NPD has a negative influence on overall programme performance and on the index of success of new products.
In our opinion, this study presents various limitations. The first is that virtually all ofthe explanatory variables were measured using a single indicator, which obviously affects their reliability and validity. However, studies available on new products have yet to suggest ways in which concepts associated with teams’ external communication can be suitably measured. Multi-sector empirical studies have also yet to be carried out.
Secondly, we believe that the possible moderating effect of the degree of product innovation on the connection between some of the explanatory variables and new product performance should be measured. In our opinion, whether or not the beneficial effect of gatekeepers (verified by previous studies) is greater in incremental innovations than in radical ones should also be calculated.
Finally, we believe that qualitative studies should be performed in order to develop adequate measures for the constructs considered in this study, as well as better understand the traits of innovative Spanish firms, which have not yet been properly identified.