Flexibility’s impact on new product development

(1)

Master’s Thesis University of Groningen Faculty of Economics and Business

Flexibility’s impact on new product

development

A meta-analysis on flexibility dimension

by

Alma Delia Sirbu

Supervisor: Hans van der Bij Coassessor: Wim Biemans

(2)

Abstract

The literature on new product development is continuously under development. Previous studies already established some very important factors that have significant impact on the product performance. One of the them is flexibility. However, the literature on product level of the flexibility is still undeveloped and incomplete. This meta-analysis tries to gather and test different meta-factors in order to comprise the homogeneous results. Moreover, after an in-depth analysis, several moderators are proposed as having a critical effect on the project outcome. Also, the study sets a starting point for future research on this matter.

(3)

1 Introduction

Einstein was once asked what he would do if he had an hour to save the world, and he responded that he would spend 55 minutes defining the problem and then five minutes solving it. His answer represents the importance of effectively spending time on problem definition.(Couger, 1995)

Although some research (e.g. Cooper & Kleinschmidt, 1987 ; Cui et al., 2011; Kettunen et al., 2015)as been carried out on the implications of adopting a more flexible approach during new product development (NPD), there have been few empirical analyses on an in-depth significance of it. This paper tries to remedy this problem by analysing the current literature with the help of meta-analysis testing.

To cope with the environmental dynamism and environmental competitiveness, the companies (e.g. Apple, Sony, IBM) nowadays are redefining their strategies. In order to handle uncertainty and risk, organizations are using flexible mechanisms, reconsidering their organizational culture and reevaluating their corporate strategies (Sushil et al., 2016). Moreover, flexibility is fundamental to knowledge creation because it facilitates the appropriation of opportunities and improves the decision-making process (Nonaka and Takeuchi, 1995).

This paper is focusing on a more in-depth approach of flexibility in NPD, more precisely, on the flexibility of defining the product during the development process. It is often considered that defining the product in an earlier stage will actually shorten the process duration. In fact, the benefits of this strategy will actually diminish over time and will lead to more time spend trying to fix the issues (Choo, 2014).

Product definition is much more than to determine the design in a certain stage of the development process. Product definition is about cooperation between different organizational units and different teams. Information sharing, cooperation and extensive communication among the teams is mandatory, in order for process and product designers to have compatible goals ( Swink, 2003).

For example, Lee and Tang (1997) introduced Delayed Product Differentiation (DPD) which is a concept to redesign the product process so the point of differentiation will be delayed in order to maximize the outcome. This also increases the flexibility of the process, enhances the service level of the system and lowers the inventory. However, it might also increase the final cost because of the redesigning costs (Lee & Tang, 1997).

(5)

informational and temporal. The organizational dimension refers to formal structuration of the entire development process; the informational dimension describes the examination of a product definition approach during the process, whereas the temporal dimension is concentrating on scheduling the tasks during stages and also the simultaneous carrying out of activities. All three perspectives will be elaborated on in the next section.

This paper empirically analyses product performance across different industries, adding important aspects to the current literature by testing the effects of multiple factors of product definition strategy and project related variables on NPD performance (on a project level). The results capture that several factors have a significant impact which contribute materially to product performance within industries.

This paper is organized as follows. First, there is an in-depth review of the literature on the flexibility’s impact on the NPD process. Second, the literature search for this study is presented along with the meta-analysis protocol. Finally, the meta-analytic testing and results of research on the information dimension of flexibility during NPD are summarized and discussed.

2 Literature Review

2.1 Flexibility during NPD process

Recent developments in flexibility within companies have led to a proliferation of studies that concentrated on different aspects of this specific domain. In product innovation management, the term flexibility is a characteristic of a process, a product, the structure of the company and the workforce (Kok & Ligthart, 2014). Prior research (Lighart, 2013; Wagner et al., 2011; Nandakumar et al., 2014; Johnson et al., 2012) has focused on different perspectives of flexibility. All of them defined, in one way or another, flexibility as a mechanism to react, adapt or anticipate changes in the internal or external environment (Wagner et al, 2011).

Researchers have tried to find out if flexibility is helping or impeding innovation in different industries (Ramirez et al., 2012; Kogut & Kulatilaka, 1994; Rainey, 2008). Another definition of flexibility can be found in the research carried out by Iansiti (1995, 1997, 2001). He defines it as the ability to embrace environmental turbulence rapidly adapting to new technological and market information that develops over the course of a project.

(6)

The organizational dimension refers to the structure of the product progression in different stages and also defining the activities that will take part of each and every stage. The organizational dimension is about planning and executing the NPD process. The organizational stages are defined by Biazzo (2009) as being “sequential temporal intervals” which can be high or low structured. The informational dimension refers to the evolution of product design and definition, analysing the degree of intersection between problem formulation and problem solving. The temporal dimension is mostly focussed on task scheduling and executing the task simultaneously by different teams. All three dimensions can be correlated and mixed for a better understanding of the extensive need of flexibility. For example when mixing informational with temporal dimension, the correlation is in fact between flexibility (informational dimension) and simultaneity (temporal dimension). This can be examined by “[…] different execution strategies: sequential, overlapped or concurrent” (Joglekar et al., 2001; Biazzo, 2009, p.346).

Biazzo (2009) also observed the perspective of the existing literature on dealing with uncertainty. First, uncertainty is defined in the literature as is the “absence of information” (Daft & Lengel, 1986). Second, there are three ways of coping with uncertainty: anticipation (Souder & Moenaert, 1992) and reaction (Verganti, 1999) and agility (Rainey, 2008). Anticipating uncertain information means to “freeze” the decisions regarding the product. The reaction strategy refers to the flexible processes which mean the concept freeze is as close as the launching stage. Also, agility means that the process is organized in a way that can be quickly changed, if necessary (Eisendhardt & Tabrizzi, 1995).

2.2 Informational Dimension of Flexibility

This study is focusing only on the informational dimension from Biazzo’s definition (2009), as there are different opposite studies in the literature, regarding early versus late definition of the product during NPD. Biazzo (2009) divided the informational dimension into two parts: problem formulation (product definition activities) and problem solving (detailed design activities). He concluded that flexibility in this dimension is not about overlapping the stages – as the other two dimensions – but rather about the design of the stages. Designing the stages is about how exactly the stages are carried out in order for problem formulation and problem solving to efficiently interact with each other. This is called organic flexibility and it presumes that high degree of interface between problem formulation and problem solving is sustained by a low structuration of the organizational dimension (Biazzo, 2009).

(7)

customer feedback,(e.g Joshi & Sharma, 2004) prototyping(e.g Verganti et al., 2010) and iterative processes(e.g Krishnan et al.,1997) play a crucial role. Customer feedback and prototyping are indirectly associated impact factors that also represent the flexibility of product design. More precise, when implementing customer feedback or building a prototype of the product, the product definition is not yet “freezed” (MacComarck et al., 2001). This means that the project is flexible and several improvements can be implemented (Eisendhardt & Tabrizzi, 1995).

2.2.1 Product definition during NPD

2.2.1 Product definition during NPD. There are different opinions about the early

definition of a product during NPD. For example, Cooper and Kleinschmidt (1994) studied whether an early sharp definition of the product has a significant impact on the speed of development. The results suggest that it is one of the top three time savers. Moreover, they stated that if the definition of the product is made during the development, and not in the pre-development stage, it might lead to delays. Furthermore, it indicates that the pre-pre-development phase was not done in a proper way, customer and market knowledge were not sufficient for an effective definition of the product (Cooper & Kleinschmidt, 1994).

Defining a product during NPD is more than solely determine some characteristics for a future product; it is about understanding customer preferences, technological risks, environment competitiveness, or emerging technologies. By this it should be understood that product definition is also about cooperation across organizational units like R&D, marketing and manufacturing. Each department(marketing, R&D, manufacturing) contributes with information and feedback regarding product features, price, and functions; all together they establish priorities during the process. They are also responsible for making any necessary changes to the product in any stage, if needed (Bacon et. al., 1994).

(8)

A good example is the paper of Krishan and Bhattacharya (1998) that created a model presenting three factors which determine the optimal product definition approach in different conditions. Those three factors are: the integration need of the specifications, the flexibility of the team, and the performance/price sensitivity of the market. They concluded that each product should be examined individually and early product definition is beneficial only in a limited number of situations; when the level of uncertainty is low, the market is price sensitive, or when the team is not flexible. In all other situations, they state that the product definition can be delayed in order to increase their profits (Krishan & Bhattacharya, 1998).

Jain and Ramdas (2005) continued the research of Krishan and Bhattacharya (1998) and found another approach to their model. Their research is based on video-gaming industry and they called it the "pace keeping approach”. They consider that especially in the gaming industry, focusing on a certain design in the beginning and having different review points for upgrading will have better outcomes. Moreover, they took into consideration the costs of changes in different points during the development in order to keep up with any alteration.

Callahan and Moretton (2001) also tested different variables and their impact on the speed of new product development in the software industry. Among the tested variables, there are “testing" and “build frequency" variables that are relevant for this study. They categorized testing in: unit tests, beta tests and system tests. These three types of testing are beneficial in the high-tech industry like software development as it is easier to discover and fix bugs, software can be tested by end users and all the data gathered along the development could be used to improve the end product/software (Callahan & Moretton, 2001).

Furthermore, they used the term "load build"; a review point of the engineers in multiple critical points of the software development. They argue that multiple load builds can have a significant impact on the NPD speed by offering quick updates and feedback to the software developers (Callahan & Moretton, 2001).

2.2.2 Custom er Feedback.

2.2.2 Customer feedback and prototyping. Customer feedback and prototyping used

(9)

information occurs during each stage of the process as a response to customer feedback.(Joshi & Sharma,2004).Moreover, Verganti et al. (2010) stated that customer feedback and prototyping are essential in dynamic environments because they have a positive effect on the speed of knowledge creation and product modification.

MacComarck et al. (2001) studied the impact of market - and technological feedback on a system level test ( where the modules are gathers in a repository) in order to improve the product to satisfy customer needs. They concluded that constructing featured prototypes, gathering feedback and overlapping stages help the product to meet the requirements of the consumers. However, different from other studies, they identified three challenges of overlapping stages. First, begin the design before the architecture of the product is finished. Second, build up the system before the actual models (requires that critical aspects of the system are done prior to the completion of the components model). The third, and most important challenge is to acknowledge the new information might be available later. Therefore, developers have to pay attention to the product architecture in order to stay flexible and to allow easy transformations (MacComarck et al., 2001).

Nevertheless, companies should take into consideration that, although an idea seems promising, it is not unusual for a product to not work as expected after building the prototype (Balachandra, 1984). This underlines the importance of several review milestones for a good development of the product.

2.2.3 Iteration

2.2.3 Iteration. Verganti et al (2010) studied the impact of an iterative process on

percentage revenue growth in relation to product performance. They emphasized the fact that in the technology-based industry, iterative operations are frequently used because of the dynamics of the environment and the close examination of new technologies on the market. They suggested that companies should consider freezing the design of the product at a later stage and to closely monitor the dynamic environment and market competition (Verganti et al., 2010).

Furthermore, Camburn et al. (2015) studied the benefits of prototyping and multiple iterations in the development process. Multiple iterations lead to a remarkable improvement of the performance up to 400% and lower fabrication time (Camburn et al., 2015).

(10)

Tabrizzi, 1995).Therefore, review points and the iteration process are effective ways of dealing with this kind of uncertainty.

Moreover, Sethi and Iqbal (2008) concluded that in a turbulent environment, early definition could be fatal for the development, as the R&D team can be trapped in a wrong perspective of the product. Krishnan and Bhattacharya (2002) had similar results and also encourage firms to stay flexible by investing in different paths from an early-stage or by "over-designing" the product even though it might be costly.

Krishnan et al. (1997) tried to demonstrate if there is an optimal number of iterations that can help giving satisfying results and minimizing the time needed for completion. They investigated the overlapping of different activities during NPD using two characteristics of the product development process, “evolution” and “sensitivity” . Ha and Porteus (1995) also tested whether there is a balance of reviews on the product definition. They state that too frequent milestones will delay the process and there would be a lack of focus on other important activities. Fewer review points will increase the probability of design flaws or ending with an unsatisfactory product (Ha & Porteus, 1995).

The goal of this paper was to check the correlation between flexibility during NPD from the informational dimension point of view and the NPD performance. First, variables were searched for in order to add value to this goal. Table 1 provides an overview of the variables used in the testing as well as their definition. It is essential to understand the focus of the study after the literature review.

Table 1

Definition of independent variables

Variables Definition Reference

Custom er F ee d b ac k m etaf ac to r 1. Information Implementation

“a change in behavior based on the knowledge that has been captured” (Lynn et al, 1999, p.444) Lynn et al (1999); Lynn et al (2000); Lynn et al (2002) 2. Customer Knowledge Development

Taking into consideration the customer new product

preferences by probing and learning activities during the development of a new product

(11)

3. Market/technologic feedback

Intensive links within the market that can be interpreted and included into the design

MacCormack et al. (2001)

4. Monitoring and feedback internal use of significant information in order to provide with an overview of the project Pinto and Prescott (1990) Ite ration M etaf a ctor 1. Instrumental Utilization Processes

How much an organization takes into consideration market information in their strategy during NPD

Moorman (1995)

2. Iteration Redesign of at least 10% of the product during the design process

Eisenhardt and Tabrizi (1995)

3. Flexible project specification

The specifications that can be changed during the

developments process (Candi et al., 2013)

Terwiesch and Loch (1998), Candi et al. (2013)

4. Freezing the product design

All the changes are allowed to a certain point in the course of development

Zirger and Hartley (1996)

5. Design change frequency How often changes are made during the process and their impact on the NPD success

(12)

6.Change High degree of newness and new information in the development process are related to the number of changes made during the process.

Stockstrom and Herstatt (2008)

7.Build frequency If innovations are associated with more frequent frames in which feedback is applied, then the team will have a greater overview of the product and its design.

Callahan and Moretton (2001)

Overall, based on the literature review of the existing literature relating product definitions during NPD to performance outcomes, it is clear that a meta-analysis research is recommended to quantitatively integrate the results found in this variety of studies.

3 Methodology

For this study, Hunter and Schimdt’s (1990) approach was used for testing. They define meta-analysis as a statistical method for cumulating results from existing literature (Hunter & Schimdt, 1990). There are several advantages of the meta-analysis approach, for instance that it is more accurate than other empirical research and allows correcting certain factors that might misrepresent important relationships (Stewart & Roth, 2003).

The meta-analysis approach is used to collect the findings across the literature to show a simplified view of relationship patterns within the studies. This type of analysis can correct for unreliability, range departure, error variance, measurement error and artifacts that can produce the “[…] illusion of conflicting findings“ (Hunter & Schimdt, 2004, pg. 17).

3.1 Literature search

(13)

different scientific databases, such as ABI/INFORM Global, ScienceDirect, Wiley Online Library, Emerald Groups, and Web of Science. The keywords and phrases used in this search were “product definition”, “new product development”, “cycle time”, “product performance”. “iteration”. “customer feedback”, “prototyping”, “design change”, and “flexible specification”. References from previous meta-analyses or scientific papers (Chen et al., 2009; Griffin 1997; Henard & Szymanski, 2001) were also taken into consideration.From 38 studies that were initially considered for inclusion in the meta-analysis, 14 papers were validated for further tests. This limitation was because I focused on some characteristics relevant for my type of study. For example, the papers had to be published in a peer-reviewed journal in order to not have biased meta-analytic results.

Second, the dependent variable had to be related to NPD performance and more importantly, focused on a project level, as opposed to a firm/organizational level. Focusing on a product rather than on a firm level improves the level of detail that the characteristics of a project are demonstrated by the characteristics of the NPD process. Further, the independent variable had to be related to product definition. In order to do the testing, the papers had to have one correlation matrix, number of respondents /sample size and Chronbach’s alpha (not mandatory) to be stated.

The dependent variable in this study is NPD performance. For NPD performance I took into account any paper that made us of speed performance, duration of NPD, product performance, financial performance or creativity performance for their dependent variable. For those studies that had more than one of the dependent variables mentioned above, a mean was calculated and used for further testing.

3.2 Meta-analysis protocol

As mentioned above, this study followed Hunter and Schimdt’s (1990) approach. Hunter and Schimdt (2004) state that the meta-analysis approach is a three step process: a set of descriptive statistics, coding the paper’s characteristics, and regression of effect sizes in the research characteristics. Or in a simplified definition, the meta-analysis approach is an analysis of analyses (Glass, 1976).

(14)

The correlation for measurement error was computed by using the respective Chronbach’s alpha from each study. If the reliability information was not presented in the correlation matrix, or in a separated table, then it was considered as being the value 1. The correlation for measurement error is the correlation for sample size divided by the square root of the reliabilities of each metafactor and of the dependent variable from each study.

Then the associated variance of ro – the total variance – was computed. This was done by using the following formula(Hunter&Schimdt, 1990):

Where Ni is the sample size of the study,

r

ooi is the correlation of each study, and

weighted rooi is the correlation for sample size, computed earlier.

(15)

Where

r

xx and

r

yy are the reliabilities of the meta-factors, respective of the

performance.

The fifth step was to compute the variance of the sampling error. The sampling error is important as it shows the variability in correlation. This variability is caused by the small sample of each study not being representative for the entire population (Hunter & Schimdt, 2004). The variance of the sampling error was calculated by squaring the correlation for sample size, subtracted from 1 and all squared again. Then, the result was divided by the average of the sample size minus 1. The formula for variance of sampling error is(Song et al.,2008):

Finally, the unexplained variance was computed by subtracting the last two computations – variance due to the artifacts and variance of sampling error – from the total variance. If this real variance was lower than 0, then it would be considered to be 0 .

In order to check whether the meta-factor is homogeneous or heterogenous, the “rule of thumb” by Hunter and Schimdt (1990) was used. This means that the unexplained variance should be less than 25% of the total variance. If it exceeds 25% of the total variance, then a moderator should be searched for or the sample is prone to error.

(16)

4 Results

After the literature review and the selection of necessary studies for a meta-analysis, the metafactors were defined: customer feedback and iteration. One of the definitions for customer feedback is presented by Joshi and Sharma (2004) as the process that helps to understand the new preferences or needs of the customers, probing and learning during several stages before commercialization with the help of the customers (Joshi & Sharma, 2004). This is highly related to this study because taking into consideration the customer feedback in a later stage means that the product definition is not yet clearly expressed.

Iteration is a variable that has plenty of different definitions. Eisenhardt and Tabrizi (1995) defined in a simple way, that iteration is the redesign of at least 10% of the product/components during NPD. They also enunciated some of the benefits of iteration during NPD like speeding the development of the product, but also for a better understanding of it. It also enables a dynamic process of decision-making by comparing different options (Eisenhardt & Tabrizi, 1995).

Table 2 Results of Meta-Analysis Metafactor K n r 95% Confidence Interval Unexplained variance Moderator Iteration 8 1100 0,1 ( -.098 - .298) 84,56% Yes Customer Feedback 6 1737 0,4139 ( .262 - .565 ) 87,56% Yes 4.1 Customer Feedback

(17)

error. Following this, three possible moderators were found: the industry, the dependent variable and the directionality of information.

Moderators analysis of customer feedback metafactor. Among the six studies from the industry point of view, four were targeting the technology industry (Lynn et al, 1999; Lynn et al, 2000; MacCormack et al, 2001; Lynn et al, 2002), one the manufacturing (Joshi & Sharma, 2004) and one several industries, but predominantly (44%) the construction industry (Pinto & Prescott, 1990). Testing the technology industry customer feedback relationship revealed inconclusive results, with r=.59 and an unexplained variance of 69,05%.

Table 3

Moderator Analysis of the Customer Feedback Metafactor

Metafactor K n Moderator r Unexplained

variance 95% Confidence interval Customer Feedback (K=6) 4 528 Technology based .41 .43 69,05% 49,24% (.31 - .5) (.37 - .48) 2 573 Non-technology based 1 29 Quality - .51 - 78,14% - (.40 - .61) 5 1072 Project performance

1 408 Internal dir. Of info. -

.51

-

81,61%

-

(.36 - .65)

5 693 External dir. of info.

The remaining studies were tested, but showed heterogeneous results, i.e. the real variance exceeded 25%; more precisely 49,25%. Thus, contrary to expectations, the industry is not a viable moderator for this study. The second moderator, the dependent variable, was then tested. Five out of six studies used the NPD Performance as the dependent variable and only one study used Quality (MacCormack et al., 2001). The results of the five studies were unsatisfactory, with a r=.41 and a real variance of 78,14%.

(18)

changes during NPD. Researchers mostly focused on external information in the decision-making process of the projects. Only the study by Pinto and Prescott (1990) specified that regarding monitoring and feedback variable, there was more emphasis on how internal information was collected and used for a better development of the project. The test was carried out on the five studies; however, the results were heterogeneous (r=.52 and unexplained variance=81,61%).

No homogeneous results were found for the relationship between the metafactor and the dependent variable. Moreover, three possible moderators were tested, but no conclusive results emerged. This might be due to the fact that the number of studies is limited (K=6) and thus, further tests should be carried out.

4.2 Iteration

Table 2 provides an overview of the results for the iteration metafactor in relation to NPD performance. It should be stated that all variables were corrected before the testing. For example,in the paper by Moorman (1995), three dependent variables were significant for this study (new product performance, timeliness, creativity) and thus, the mean of these variables was used in the calculation. Further, there was a difference between the studies in approaching the time variable. Some of the researchers (e.g. Souder et al., 1998) focused on the time performance as their dependent variable, measuring the time needed for a project, and not if the project was on time. The variables were inverted in order to be used in this study. The remaining studies focused on project performance as dependent variables.

As visible in Table 2, no homogeneous results were found in the first testing. The correction for sample size is very low (-.04) and the unexplained variance is almost the same as for the customer feedback metafactor (approximately 85%).

Moderator Analysis of the Iteration Metafactor. For this metafactor, two possible moderators were found: the industry and the dependent variable. Mostly, the studies gather data from the technology industry as it is a more complex industry compared to non-technology based industries.In the non-technology indsutry projects are full of uncertainty and more difficult to develop successfully (Lynn, 1999).

The results from the testing iteration with the first moderator – the industry –were not homogeneous, with a low correlation for the sample size of .07 and a high unexplained variance of 87,32%.

(19)

2013). The explained variance was 0.; thus, there is a significant correlation between iteration and non-technology industries.

Table 4

Moderator Analysis of the Iteration Metafactor

Metafactor K n Moderator r Unexplained

variance 95% Confidence interval Iteration (K=8) 6 876 Technology based .07 .21 87,32% 0 (-.14 - .28) (.21 - -.21) 2 224 Non-technology based 6 493 Time performance .15 .05 87,24% 23,91% (-.13 - .43) (-.2 - .3) 2 607 Project performance

Furthermore, the results from testing the iteration impact on project performance with the dependent variable as a moderator were also satisfactory. The findings from the studies that used time performance as a dependent variable did not conclude with the necessary results, with a real variance 87% of total variance.

Testing the other two studies (Candi et al., 2013; Stockstrom & Herstatt, 2008) that were isolated because of the moderator, resulted with homogeneous results with a real variance of <25%. Actually, both studies belong to the same dependent variable – project performance – and this act as a moderator for the iteration’s impact on the project performance.

(20)

5 Discussion and Conclusion

This study tries to demonstrate that there are several factors that have a significant impact on the progress of a product. One of the main factors is flexibility, and more precisely, the flexibility in defining the product during the NPD. In a highly dynamic environment, increasing the flexibility and decreasing the rigidity of the process is one of the main approaches that researchers are making use of. This study adds significant results to the literature of flexibility.

Biazzo (2009) analyses a different perspective of flexibility, comprising three dimensions: the organizational, informational and temporal dimensions. As these three dimensions are ample and vast factors, they can be analyzed individually. This study sought to determine the impact of one dimension on product performance: the informational dimension.

Two moderators out of five showed homogeneous results. This inconsistency can be explained by the contradictory studies that were used for testing. Moreover, there might be other moderators that could present homogeneous results, but the literature is not currently focusing in this direction.

This study offers new insights regarding the association between product flexibility and product performance. It indicates that there is growing evidence that some moderators have a significant impact on freezing the product design at a certain time.

In conclusion, this study illustrates the theoretical and practical value of a product definition. It adds to the current literature by introducing a new perspective to the management of new product development. However, it should be noted that several organizations still have a myopic view on the importance of product design and its outcomes. These results compel organizational strategies towards a more flexible approach of new product development.

5.1 Managerial Implications

This paper addresses an important question by managers: Is early definition of a product a good strategy? The findings suggest that managers should focus more on expanding their flexibility strategy during the development process than using more rigid processes.This strategy might shorten the development time and increase product performance.

(21)

done at a small scale , it is offering new insights on the literature and a starting point for future research.

In addition, this study increases the level of understanding of managers and researchers because of the in-depth literature review and the meta-analytic tests in the flexibility domain.

5.2 Limitations

The limitations of this study are mostly coming from a lack of literature on this subject. Most papers used for this meta-analysis are approaching only a broad view of the flexibility dimension. The current literature is still not mature and future research should be done more deeply into these three dimensions: informational, temporal and organizational. Because of that, the limitation of this study is that the sample might not be accurate because of the small sample size.

Also, another limitation is the absence of neccesary correlation variables in various studies that could have helped in obtaining more homogeneous results. Only fourteen studies out of thirty-eight actually had almost all the variables needed for testing ( some of them didn’t present the Chronbach’s Alpha).

5.3 Future Research

No significant results of the meta-analysis were found. However, there are two moderators of the metafactors that can be a starting point for future research. Future research could consider the informational dimension in order to expose more relevant results. Concerning the literature search, few empirical papers are approaching a more comprehensive concept on the advantages and disadvantages of freezing the product definition during the NPD. More research could investigate this field as it is an important factor that can increase product performance. Moreover, more researchers are nowadays focusing on customer feedback implementation in the stages of the NPD, but there are few papers that are also taking into account the indirect consequences of this factor. Future research should also be carried out on this issue. This could prove to be an important issue for future research; especially finding and testing other moderators is vital in this sense. Surprisingly, the non-technology based industries acted as a significant moderator for this meta-analysis. This means that future research might approach this moderator as a starting point.

(22)

dynamism and much more. It should be analyzed how internal and external factors could have a significant impact on the product definition during the development in order to increase the product performance.

(23)

References

Bacon, G., Beckman, S., Mowery, D., & Wilson, E. (1994). Managing product definition in high-technology industries: a pilot study. California Management Rev,. 36, 32–56. Balachandra, R. (1984), Critical Signals for Making Go/NoGo Decisions in New Product

Development. Journal of Product Innovation Management, 1, 92–100.

Bardhan, I.R., Krishnan, V., Lin, A. (2013) Team dispersion, information technology, and project performance. Production and Operation Management, 22(6), 1478–1493 Bhattacharya, S., Krishnan, V., & Mahajan, V. (1998). Managing New Product Definition in

Highly Dynamic Environments. Management Science, 44(11-part-2), 50-S64.

Biazzo, S. (2009), Flexibility, Structuration, and Simultaneity in New Product Development. Journal of Product Innovation Management, 26, 336–353.

Callahan, J. & Moretton, B. (2001) Reducing Software Product Development Time. International Journal of Project Management, 19(1), 59-70.

Candi, M., van den Ende, J., & Gemser, G. (2013). Organizing Innovation Projects under Technological Turbulence. Technovation, 33, 133-141.

Cankurtaran, P., Langerak, F., & Griffin, A. (2013), Consequences of New Product Development Speed: A Meta-Analysis. J Prod Innov Manag, 30, 465–486.

Chen, J., Damanpour, F., & Reilly, R. R. (2010). Understanding antecedents of new product development speed: A meta-analysis. Journal Of Operations Management, 28(1), 17-33.

Choo, A. S. (2014). Defining Problems Fast and Slow: The U-shaped Effect of Problem Definition Time on Project Duration. Production And Operations Management, 23(8), 1462-1479. doi:10.1111/poms.12219

Cooper, R. G., & Kleinschmidt, E. J. (1987). New products: What separates winners from losers? The Journal Of Product Innovation Management, 4(3), 169-184. doi:10.1016/0737-6782(87)90002-6

Cooper, R.G. & Kleinschmidt, E. (1994), Determinants of timeliness in product development. Journal of Product Innovation Management, 11, 381-96.

(24)

Crawford, J. C. (2007) Product Innovation: Leading Change Through Integrated Product Development By David L. Rainey. Journal of Product Innovation Management, 24(2), 188-190

Cui A.S., Zhao M., & Ravichandran T. (2011). Market uncertainty and dynamic new product launch strategies: A system dynamics model. IEEE Transactions On Engineering Management, 58(3), 530-550. doi:10.1109/TEM.2010.2100822

Eisenhardt, K.M. & Tabrizi, B.N. (1995) Accelerating Adaptive Processes: Product Innovation in the Global Computer Industry. Administrative Science Quarterly, 40, 84- 110.

Fuchs, C. & Schreier, M. (2011), Customer Empowerment in New Product Development*. Journal of Product Innovation Management, 28, 17–32.

Griffin, A. (1997). Modeling and measuring product development cycle time across industries. Journal Of Engineering And Technology Management, 14(1), 1-24.

Ha, A. Y., E. L. Porteus. 1995. Optimal timing of reviews in concurrent design for manufacturability. Management Science 41(9) 1431-1447

Hunter, J.E. & Schmidt, F.L. (1990). Methods of Meta-Analysis: Correcting Error and Bias in Research Findings. Newbury Park, CA: Sage.

Hunter, J.E. & Schmidt, F.L. (2004). Methods of Meta-Analysis: Correcting Error and Bias in Research Findings, 2nd ed. Thousand Oaks, CA: Sage.

Iansiti, M. (1995). Shooting the Rapids: Managing Product Development in Turbulent Environments. California Management Review, 38(1), 37–58.

Iansiti, M. (1997). From technological potential to product performance: an empirical analysis. Research Policy, 26(3), 345-365.

Jain, S. & Ramdas, K. (2005). Up or Out—or Stay Put? Product Positioning in an Evolving Technology Environment. Production and Operations Management, 14, 362–376. Johnston, R., Clark, G., & Shulver, M. (2012). Service Operations Management. 4th Edition.

Essex: Prentice Hall.

Joshi, A.W., &, Sharma, S. (2004). Customer Knowledge Development: Antecedents and Impact on New Product Performance. Journal of Marketing, 68(4), 47-59.

Kalyanaram, G., & V. Krishnan. (1997). Deliberate product definition: customizing the product definition process. J. Marketing Res, 34, 276–285.

(25)

Operational Research 244(3), 892-904

Kok, R. A. W. & Ligthart, P. E. M. (2014). Differentiating Major and Incremental New Product Development: The Effects of Functional and Numerical Workforce Flexibility*. J Prod Innov Manag, 31, 30–42.

Kogut, B., & Kulatilaka, N. (1994). Operating Flexibility, Global Manufacturing, and the Option Value of a Multinational Network. Management Science, 40(1), 123-139. Krishnan, V. & S. Bhattacharya (2002). Technology Selection and Commitment in New

Product Development: The Role of Uncertainty and Design Flexibility. Management Science, 48, 313-327.

Lafou, M., Mathieu, L., Pois, S., & Alochet, M. (2016). Manufacturing System Flexibility: Product Flexibility Assessment. Procedia CIRP, 41, 99-104.

Lee, H. L., & Tang, C. S. (1997). Modelling the costs and benefits of delayed product differentiation. Management science, 43(1), 40-53.

Ligthart R., Oerlemans L., & Noorderhaven N. (2016). In the Shadows of Time: A Case Study of Flexibility Behaviors in an Interorganizational Project. Organization Studies, 37(12), 1721-1743.

Lynn, G., Richard, Skov, B., & Abel, K. (1990). Practices that Support Team Learning and Their Impact on Speed to Market and New Product Success. Journal of Product Innovation Management, 16(5), 439-454

Lynn, G., Reilly, R., & Akgun, A. (2000). Knowledge Management in New Product Teams: Practices and Outcomes. IEEE Transactions on Engineering Management, 47(2), 221- 232.

Lynn, G., Reilly, R., & Akgun, A (2002). Multi-Dimensionality of Learning in New Product Development Teams. European Journal of Innovation Management, 5, 57-72

MacCormack, A., Verganti, R., & Iansiti, M. (2001). Developing Products on "Internet Time": The Anatomy of a Flexible Development Process. Management Science, 47(1), 133-150.

MacKelprang, A. W., & Nair, A. (2010). Relationship between just-in-time manufacturing practices and performance: A meta-analytic investigation. Journal of Operations Management, 28(4), 283-302.

(26)

Nandakumar, M. K., Jharkharia, S., & Nair, A. S. (Eds.). (2014). Organisational flexibility and competitiveness. Springer.

Nonaka, I. and Takeuchi, H. (1995),The Knowledge-creating Company: How Japanese CompaniesCreate the Dynamics of Innovation, Oxford University Press, Oxford. Camburn, B. A., Jensen, D., Crawford, R., Otto, K., & Wood, K. (2015). EVALUATION OF

A STRATEGIC METHOD TO IMPROVE PROTOTYPE PERFORMANCE WITH REDUCED COST AND FABRICATION TIME. In DS 80-4 Proceedings of the 20th International Conference on Engineering Design (ICED 15) Vol 4: Design for X, Design to X, Milan, Italy, 27-30.07. 15.

Pinto, J., & Prescott, J. (1990). Planning and Tactical Factors in the Project Implementation Process. Journal of Management Studies, 27(3), 305-327.

Mihi Ramirez, A., Jesús García Morales, V., & Arias Aranda, D. (2012). Knowledge creation and flexibility of distribution of information. Industrial Management & Data Systems, 112(2), 166-185.

Sethi, R., & Iqbal, Z. (2008). Stage-gate controls, learning failure, and adverse effect on novel new products. Journal of Marketing, 72(1), 118-134.

Song, M., Podoynitsyna, K., Van Der Bij, H., & Halman, J. I. M. (2008). Success Factors in New Ventures: A Meta-analysis<sup>*</sup>. Journal Of Product Innovation Management, 25(1), 7-27. doi:10.1111/j.1540-5885.2007.00280.x

Souder, W. E., & Moenaert, R. K. (1992). Integrating marketing and R&D project personnel within innovation projects: an information uncertainty model. Journal of management studies, 29(4), 485-512.

Souder, W., Sherman, D., & Davies-Cooper, R. (1998). Environmental Uncertainty, Organisational Integrations, and New Product Development Effectiveness: A Test of Contingency Theory. Journal of Product Innovation Management, 15, 520-533.

Stockstrom, C., & Herstatt, C. (2008). Planning and uncertainty in new product development. R&d Management, 38(5), 480-490.

Sushil, Bhal, K. T., & Singh, S. P. (2016). Managing flexibility : People, process, technology and business (Flexible systems management; Flexible systems management series). New Delhi: Springer.

Swink, M. (2003). Completing projects on-time: how project acceleration affects new product development. Journal Of Engineering And Technology Management, 20(4), 319-344. doi:10.1016/j.jengtecman.2003.08.002

(27)

Projects: Balancing Firmness and Flexibility in the Innovation Process. Journal of Operations Management, 18, 401–425.

Terwiesch, C., & Loch, C. (1999). Measuring the Effectiveness of Overlapping Development Activities. Management Science, 45(4), 455-465.

Thomke, S.H. (1997). The Role of Flexibility in the Development of New Products: An Empirical Study. Research Policy 26(1): 105–119.

Van Oorschot, K., Sengupta, K., Akkermans, H. and Van Wassenhove, L. (2010). Get Fat Fast: Surviving Stage-Gate® in NPD. Journal of Product Innovation Management, 27, 828–839.

Verganti, R. (1999). Planned flexibility: linking anticipation and reaction in product development projects. Journal of Product Innovation Management, 16(4), 363-376. Buganza, T., Gerst, M., & Verganti, R. (2010). Adoption of NPD flexibility practices in new

technology-based firms. European Journal of Innovation Management, 13(1), 62-80. Wagner, D., Suchan, C., Leunig, B., & Frank, J. (2011, February). Towards the analysis of

information systems flexibility: proposition of a method. In Proc 10th international conference on Wirtschaftsinformatik, Zurich (pp. 808-817).

(28)

Appendix

Appendix A1-Customer Feedback

Appendix A2-Customer Feedback- technology industry moderator

Appendix A3-Customer Feedback- non-technology moderator

Appendix A4-Customer Feedback-dependent variable moderator

(29)

Appendix A5-Customer Feedback=directionality of information moderator

Appendix A6-Iteration

Appendix A7-Iteration - technology industry moderator

(30)

Appendix A9-Iteration - time performance moderator

Flexibility’s impact on new product development