Organizing innovation projects under technological turbulence

Technovation 33 (2013) 133–141

Contents lists available at SciVerse ScienceDirect

Technovation journal homepage: www.elsevier.com/locate/technovation

Organizing innovation projects under technological turbulence Marina Candi a,n, Jan van den Ende b,1, Gerda Gemser c a b c

Reykjavik University, School of Business, Menntavegur 1, 101 Reykjavik, Iceland Rotterdam School of Management, Department of Management of Technology and Innovation, P.O. Box 1738, 3000 DR Rotterdam, Netherlands School of Economics, Finance and Marketing, RMIT University, GPO Box 2476V, Melbourne 3001, Australia

a r t i c l e i n f o

abstract

Available online 16 February 2013

Research on the organization of innovation projects suggests that increased project flexibility is a common reaction to high levels of technological turbulence. However, existing definitions of project flexibility are inconsistent and sometimes unclear, and empirical evidence is limited. This article makes an important distinction between flexible project planning and flexible project specifications. A negative relationship is found between flexible project planning and innovation project performance, whereas flexible product specifications are found to contribute positively. This article also examines how technological turbulence contributes to the choice of flexible or inflexible strategies. Technological turbulence can be present in the external environment or can be internal to the firm, when radically new products are developed. The findings suggest that when businesses perceive technological turbulence in the environment they are more likely to adopt flexible approaches to innovation in an attempt to adapt to external pressures. In technologically innovative projects, product specifications are likely to remain fixed while project organization is likely to be adapted to the needs of the project. Taken together, the findings suggest that innovation projects should maintain stable organization, schedules and budgets, but stay flexible about product specifications. Vigilance with regards to external and internal conditions of technological turbulence, which may lead organizations to be more flexible in terms of project planning, is needed. & 2013 Elsevier Ltd. All rights reserved.

Keywords: Innovation project organization Flexible project planning Flexible specifications Technology turbulence

1. Introduction Environmental dynamics need to be taken into account in the management of innovation (Drejer, 2002), particularly turbulence (Ferna´ndez et al., 2010; Lee and Wong, 2011). Project flexibility has been proposed as a means to cope with high levels of turbulence (Eisenhardt and Tabrizi, 1995; MacCormack et al., 2001; Garud et al., 2008; Levardy and Browning, 2009; Moorman and Miner, 1998; Lenfle and Loch, 2010). However, definitions and operationalizations of project flexibility are varied and sometimes unclear. Research arguing for flexible approaches mostly falls broadly into two camps with regard to what is meant by flexibility. On one hand, there is flexible project planning, which may include overlapping development phases (Iansiti and MacCormack, 1997), improvisation (Moorman and Miner, 1998), short milestones (Eisenhardt and Tabrizi, 1995), crossfunctional teams (Buganza et al., 2009), flexible use of resources (Li et al., 2010) and adaptive processes (Levardy and Browning, 2009). On the other hand, there are flexible project specifications, which can entail practices such as trial-and-error iteration or prototyping (Eisenhardt n

Corresponding author. Tel.: þ354 599 6200. E-mail addresses: [email protected] (M. Candi), [email protected] (J.v.d. Ende), [email protected] (G. Gemser). 1 Tel.: þ31 10 4082299. 0166-4972/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.technovation.2013.01.002

and Tabrizi, 1995; Souder et al., 1998; Buganza et al., 2009; Lenfle and Loch, 2010) and postponed concept freeze (Iansiti and MacCormack, 1997; Buganza et al., 2009). There is also research that defines project flexibility in terms both of practices relating to project planning and to product specifications (e.g. MacCormack et al., 2001; Eisenhardt and Tabrizi, 1995; Buganza et al., 2009). Because of these different and sometimes overlapping conceptualizations of project flexibility, it can be difficult to gain a holistic understanding of performance implications. Although the theme of technological turbulence is widely included in research on flexible project organization, the source of the turbulence that might contribute to—or call for—the adoption of flexible strategies can differ. Existing literature has focused on technological turbulence resulting from the external environment (e.g. Buganza et al., 2009; Moorman and Miner, 1998) or the turbulence resulting from the decision to pursue innovation projects with a high degree of technological novelty (e.g. Tatikonda and Montoya-Weiss, 2001; Eisenhardt and Tabrizi, 1995). Even though existing research suggests that project flexibility is more common under conditions of technological turbulence (Buganza et al., 2009; Moorman and Miner, 1998), it is not clear to what extent this applies for turbulence caused by the external environment or by turbulence caused by the type of innovation project at hand.

134

M. Candi et al. / Technovation 33 (2013) 133–141

This paper makes an important contribution to research on the organization of innovation projects by distinguishing explicitly between two types of flexibility: flexible project planning and flexible product specifications. As discussed above, both have been included in existing definitions of flexibility, but we posit that they are conceptually different, will be chosen under different conditions and have different effects on performance. Flexible project planning refers to flexibility in project organization, scheduling and budget, whereas flexible product specifications refer to flexibility in the definition of the product to be developed. The reason for making this distinction is the expectation that planning and specifications address different aspects of project organization, with potentially different performance effects. Not only technological turbulence existing in the environment but also the newness of technology to be developed has been identified as an important source of uncertainty in product development (Tatikonda and Montoya-Weiss, 2001). Therefore, we also make an important distinction between external technological turbulence present in a firm’s environment and internal technological turbulence due to technological innovation. We examine how both types of technological turbulence contribute to the choice of both types of flexibility. In addition to technology turbulence, uncertainty may also be caused by market turbulence (Calantone et al., 2003). However, our focus is on technological turbulence rather than market turbulence since the implications of technological turbulence for innovation project organization have been found to be more important than market turbulence (Sethi and Iqbal, 2008). Our findings suggest that when businesses perceive technological turbulence in the environment they are likely to attempt to adapt to these external pressures by adopting more flexible project planning and more flexible project specifications. Conversely, a high degree of technological innovativeness, which is likely to result in internal technological turbulence, is found to contribute negatively to flexible product specifications. Like external technological turbulence, technological innovativeness contributes positively to flexible project planning. This suggests that in technologically radical innovation projects, product specifications are likely to remain fixed while project organization is likely to be adapted to the needs of the project. Furthermore, our findings suggest that flexible project planning contributes negatively to project performance, whereas flexible product specifications have a positive effect. A possible explanation is that flexible product specifications can facilitate the adaptation of a new product to market conditions whereas flexible project planning may lead to schedule delays and budget overruns, which in turn may lead to higher prices, later-than-competitor market introduction, and thus to lower project performance. Together these findings suggest important practitioner implications. Practitioners organizing innovation projects should take care to distinguish between the different types of flexibility, and give their teams clear timing and budget targets, but stay flexible about product specifications. They should be especially vigilant about external and internal conditions of technological turbulence that may sway their strategy to be more flexible in terms of project planning, with potentially detrimental effects on project performance.

2. Background and hypothesis development Methods for organizing innovation projects (also commonly referred to as new product development projects) espoused by the existing literature fall into three categories: sequential methods (Cooper, 2011), parallel or concurrent methods (Wheelwright and Clark, 1992) and flexible methods (Eisenhardt and Tabrizi, 1995; MacCormack et al., 2001).

Sequential approaches divide product development into a set of tasks undertaken one after another, resulting in projects that are easy to track and manage. A well-known and popular sequential approach to innovation is the stage gate model (Cooper, 1994, 2009). However, despite claims about the flexibility of sequential models (Cooper, 2009), in practice they tend to suffer from detrimental rigidity, particularly under conditions of technological turbulence (Sethi and Iqbal, 2008; Iansiti and MacCormack, 1997; Lenfle and Loch, 2010; Bhattacharya et al., 1998). The parallel model of product development, also referred to as concurrent engineering, is characterized by overlapping phases. The essence of this model is that downstream activities, such as the development of a manufacturing unit or marketing plan, are started during earlier stages, particularly during the design stage (Wheelwright and Clark, 1992; Krishnan et al., 1997). Advantages attributed to the parallel model are improved communication between functional activities (Wheelwright and Clark, 1992), which in the end can save time and costs, while enhancing quality. However, existing research has been inconclusive with regards to the effectiveness of concurrent engineering (Tatikonda and Montoya-Weiss, 2001; Iansiti, 1995). A third category of innovation organization strategies, and the one on which this paper focuses, is a flexible approach. Eisenhardt and Tabrizi (1995) compare a relatively inflexible parallel approach (compression model) with a flexible approach (experiential model) and argue that the parallel approach is more suitable for stable circumstances and the flexible one for dynamic environments. Subsequent research has suggested that a flexible project approach is particularly effective in turbulent contexts (e.g., MacCormack et al., 2001; Moorman and Miner, 1998). However, although the work of Eisenhardt and Tabrizi (1995) has been highly cited, it has not resulted in a clear definition of what a flexible approach involves. Eisenhardt and Tabrizi (1995) define flexibility as an approach to cope with uncertainty in shifting markets and technologies and operationalize it in terms of multiple design iterations, extensive testing, frequent milestones to assess project progress, and powerful project leadership. Using what they refer to as the ‘high-velocity’ computer industry as their empirical setting, Eisenhardt and Tabrizi (1995) find that adoption of these tactics can substantially shorten product development times. MacCormack et al. (2001) propose a definition of flexibility that includes overlapping phases, team adaptability, delay of concept freeze and regular testing. They position their research in the Internet software development industry, which they posit is an industry characterized by high levels of uncertainty and technological turbulence. They find that a more flexible development process is associated with better performing projects in terms of product quality. Buganza et al. (2009) operationalize flexibility in terms of rapid iterations, delay of concept freeze, early experiments involving customers, as well as cross-functional teams and a flat organizational structure. According to Buganza et al. (2009) particularly the last two of these strategies support accelerated project iterations and thus support practices that result in flexible product specifications. Buganza et al. (2009) do not assess performance implications of these strategies, but based on case study data they find that firms tend to adopt delay of concept freeze and rapid project iterations—but not the other strategies identified—when there is a high level of technological turbulence. Based on a case study of the management of an innovative project in the defense industry (the ‘Manhattan Project’), Lenfle and Loch (2010) draw lessons about the need for trial-and-error iterations and prototyping—rather than phased, sequential project organization—to come up with an effective solution that may not have been envisioned at the outset. In contrast with the

M. Candi et al. / Technovation 33 (2013) 133–141

135

research discussed earlier, Lenfle and Loch (2010) are thus espousing flexibility specifically in terms of flexible product specifications even though, similarly to the assertions of Buganza et al. (2009), they also link flexible processes with flexible specifications (cf. Thomke and Reinertsen, 1998). Two potential explanations for the seeming inconsistency of existing research on flexibility in innovation projects, particularly under conditions of technological turbulence, are noted from the review of existing research above. Firstly, some researchers tend to focus on flexible project planning, others focus on flexible product specifications and still others do not make an explicit distinction between the two. Eisenhardt and Tabrizi (1995) operationalize flexibility in terms of multiple design iterations, extensive testing, frequent milestones to assess project progress, and powerful project leadership. Thus, their operationalization can in fact be said to encompass a combination of flexible project specifications (multiple design iterations) and inflexible project planning (extensive testing, frequent milestones). Moorman and Miner (1998) operationalize their concept of flexibility (referred to by them as improvisation) as resulting when the definition of a plan of action and its execution overlap. This suggests they view flexibility as a combination of flexible project specifications and flexible project planning. As discussed earlier, our research makes the important distinction between these two types of flexibility in order to better understand how each contributes to project performance. Secondly, although the theme of turbulence, dynamism or uncertainty is widely included in existing research, some researchers focus on the turbulence present in the environment at each given time, while others focus on the internal turbulence that goes hand-in-hand with radical innovation.

time spent on the project short and costs low. Therefore, while flexibility in project planning may facilitate the adaptation of project contents to evolving market requirements, it may not necessarily lead to improved performance outcomes. Project teams and managers are often confronted with trade-offs between, for instance, improving the quality of a product further at the expense of schedule delays and budget over-runs (Blindenbach-Driessen et al., 2010). For example, managers have been reported to deliberately take products out of the hands of innovation teams engaged in refining and debugging to be able to launch them on time (Kawasaki, 1999). Clear deadlines with respect to timing and budget limits are likely to encourage innovation teams to develop a product within the time and cost constraints set (Langerak and Hultink, 2006). As a consequence, they will be stimulated to adapt the product to defined customer requirements in a timely way. A high degree of project planning flexibility will postpone the need to make such adaptations, and thus may have the effect that such adaptations are performed late in the process, if at all. Indeed, a major criticism of the flexible approach is the propensity to continue to wait for new information, which can result in severe delays (Pina e Cunha and Gomes, ¨ 2003). Brattstrom et al. (2012) suggest that a flexible project planning may have a detrimental effect on process outcomes such as trust and creativity, which in turn may result in poorer performance. In the same vein, Olausson and Berggren (2010) argue that it is advisable to balance flexible product specifications with structured project planning. Thus, we hypothesize for a negative contribution of flexible project planning.

2.1. Flexible project specifications

2.3. External technology turbulence and flexibility

The ability to adapt product specifications to changing requirements in the environment has been identified as an important strategy for success (Knot et al., 2001; Dickson, 1997; Lenfle and Loch, 2010; Thomke and Reinertsen, 1998; Thomke, 1997; Schlesinger et al., 2012) particularly for radical innovation (Bessant et al., 2010). Practices that support flexibility in product specifications to deal with uncertain environments are sometimes referred to as agile product development (Thomke and Reinertsen, 1998). MacCormack et al. (2001) find positive effects of delay of concept freeze on project quality. Likewise, based on case research conducted in the software industry, Iansiti and MacCormack (1997) find an overall negative effect on product quality of having a high percentage of functionality defined early. This leads them to conclude that a flexible approach to product specifications is advisable. In earlier work, Iansiti (1995) found that delay in concept freeze enhanced system-level performance of innovation projects in the technologically turbulent mainframe and supercomputer industry. In the same vein, Thomke (1997) found that, in the integrated circuit design industry, innovation projects using flexible designs that can be modified relatively quickly and at low cost in reaction to turbulence, outperformed projects employing inflexible designs. We therefore hypothesize that flexible product specifications will contribute positively to new product performance.

Our second focus of examination is on the conditions under which the two types of flexibility are likely to be chosen. As outlined above, much of the literature reviewed has emphasized the need for flexibility in innovation projects in turbulent environments. A firm’s external environment can become turbulent when a new technology emerges leading to new market expectations or rapid changes in technology (Lee and Wong, 2011). Bessant et al. (2005) refer to the need to manage ‘‘beyond the steady state’’ (p. 1366), meaning that management must adapt to external turbulence. MacCormack et al. (2001) argue for the need for project flexibility in turbulent environments, but their choice of a turbulent empirical setting does not allow for comparisons with less turbulent contexts. Nor do they explicitly distinguish between internal turbulence and environmental turbulence, although it is implied that both are high in their chosen industry context. Buganza et al. (2009) examine the application of a flexible approach in nine service development projects in firms in the telecommunication industry and find that flexible strategies are more commonly applied under conditions of high environmental turbulence. In a similar vein, Moorman and Miner (1998) find that what they refer to as improvisation (defined as happening when the definition of a plan of action and its execution overlap), which is one variant of flexible project planning, is more common under conditions of high environmental turbulence. Based on the research discussed above, we expect that flexibility will be more likely to be the strategy of choice in technologically turbulent environments than in stable environments.

H1. Flexible product specifications contribute positively to new product performance.

H2. Flexible project planning contributes negatively to new product performance.

2.2. Flexible project planning

H3. External technological turbulence contributes positively to the adoption of a flexible strategy with regards to product specifications.

Flexibility in project planning may reduce the urgency for innovation teams to work efficiently and effectively and keep the

H4. External technological turbulence contributes positively to the adoption of a flexible strategy with regards to project planning.

136

M. Candi et al. / Technovation 33 (2013) 133–141

2.4. Internal technology turbulence and flexibility

3. Methodology

Internal technology turbulence is likely to result within a business when a decision is made to develop a new technology. Eisenhardt and Tabrizi (1995) conceptualize turbulence as the relative certainty of an innovation project ranging from uncertain (high turbulence) to certain (low turbulence). Tatikonda and Montoya-Weiss (2001) define turbulence similarly. While we hypothesized that external technological turbulence is likely to lead to more flexible product specifications, we hypothesize that technological innovativeness, being controlled internally, is likely to lead to more fixed product specifications. This is based on the notion that firms will seek anchoring in a set of product specifications when internal turbulence is high (Kawasaki, 1999).

3.1. Sample and data collection

H5. Technological innovativeness contributes negatively to the adoption of a flexible strategy with regards to product specifications. Tatikonda and Montoya-Weiss (2001) examine flexibility in terms of process adaptability defined as the degree of discretion available to project managers to adjust work activities and decisions during the project and find that the technological novelty of a project increases the positive contribution of process adaptability to project outcomes. Veryzer (1998) examines how the organization of radically innovative projects differs from that of more incremental projects. Based on case research in eight projects, he concludes that projects involving radical innovations do not follow conventional organization, e.g. stage-gate, but rather are characterized by phase overlap and informal management. Thus, these findings suggest that when faced with internal technological turbulence stemming from radical innovation, firms are likely to adopt more flexible project planning.

H6. Technological innovativeness contributes positively to the adoption of a flexible strategy with regards to project planning. The hypotheses to be tested are shown in Fig. 1. In line with the contingency model proposed by Cardinal et al. (2011), we expect that the environment will influence a firm’s level and type of project flexibility, which in turn will contribute to performance. Cardinal et al. (2011) depart from the more common view in much extant innovation literature that performance is a function of interaction contingency between the environment and project organization. Based on field study observations and simulation results, they suggest that the technological environment shapes performance priorities (e.g. meeting cost or time-to-market targets), which in turn influences project management, and ultimately performance outcomes.

External technology turbulence

H3+

The hypotheses were tested using data collected about innovation projects in 132 Dutch firms. In the second half of 2009, a list of almost 700 firms belonging to a wide range of industry sectors was generated using a random process. This made up the initial pool of potential participants. Before contacting each firm, its website was examined for evidence that the firm had launched a new product within the year 2008. The reason for this specific time frame was to maximize the likelihood that respondents would remember the innovation project in question well enough to be able to respond to queries about the process involved, while also increasing the likelihood that there would already be evidence of new product performance. If there was no evidence of new product launch that could possibly fall within the defined time frame on a firm’s website, no further action was taken for that firm. Managers of about 500 firms were contacted by telephone and first asked screening questions to confirm at least one new product launch in the defined time frame. If there was no such product launch, participation was not requested. If the basic criteria for inclusion were met, a request was made for survey participation by two managers: the project manager who had been responsible for the innovation project previously identified and a business manager who could provide general information about the firm, its strategies, competitive environment and performance of the new product. Of the 500 firms contacted 163 (33%) agreed to participate in the survey. The manager initially contacted by phone was asked to name the two respondents, which frequently included him or her. Each of the potential respondents was contacted by phone and an appointment was made to conduct a telephone survey using one of two questionnaires. The first questionnaire was for project managers and covered the development of a specific new product and its characteristics. The second questionnaire was for business managers and focused on market success of the new product and firm-level variables. By collecting data for the independent and dependent variables from different respondents, the potential problem of common method bias was greatly mitigated. As was to be expected, not all the named respondents were willing to participate, and the result was 132 matched pairs of business manager and project manager responses, which were used to test the hypotheses. The firms included in the sample were drawn from a wide range of industry sectors, as shown in Table 1. The firms ranged in size from fewer than 10 employees (13% of the sample) to more than 1000 employees (13% of the sample) and the median size was 140 employees. Average values of variables were compared across the categories of sectors and sizes of firms and were not found to differ substantially between

Flexible specifications H1+

H4+ Project performance H5÷ Technological innovativeness

H6+

Flexible planning

Fig. 1. Research model.

H2÷

M. Candi et al. / Technovation 33 (2013) 133–141

groups. The innovation projects studied spanned a broad range of types as was to be expected given the wide range of sectors included. Some examples of innovation projects studied include lighting systems, food products, cosmetics, building systems, laboratory instruments, sport products, watering systems, hotels, financial services, environmental sustainability products, medical equipment, and many more. The surveys were developed in English and then translated into Dutch by a professional translator. This was followed by translation back to English by a second professional translator. Discrepancies that came up in the back-translation process were re-translated and verified until we were certain that both language versions were equivalent. 3.2. Measurement model Information about the composition of each of the dependent and independent variables is shown in the Appendix. This includes the texts of the items, response ranges, respondents, references for the items and reliability estimates. Formative indicators are appropriate when the composition of variables is driven by conceptual criteria rather than by representativeness as is the case with the more traditionally used reflective variables (Diamantopoulos and Winklhofer, 2001). The items included in our variables for flexible product specifications and flexible product planning are additive in nature, meaning that each one adds something separate to the variable. Thus, it was deemed appropriate to use formative indicators. Both formative indicators were verified using principal component analysis and variance inflation factors (see Appendix) were checked to confirm the absence of unacceptable collinearity between the items. A formative indicator of flexible product specifications was made up of questions about the degree to which the requirements for the product or service were defined in the project brief (Iansiti and MacCormack, 1997), the degree to which management allowed modifications of the project brief during the project (Eisenhardt and Tabrizi, 1995; Souder et al., 1998; Buganza et al., 2009; Lenfle and Loch, 2010) and the degree to which the project concept was frozen

Table 1 Categories of sectors represented in the sample. N ¼132. Sector

Proportion of sample

Food products, apparel, furniture and fixtures and kindred products Paper, chemicals, rubber and miscellaneous plastics products Metal, stone, clay, glass and concrete products Industrial and commercial machinery, transportation equipment, computer equipment Electronics, measuring, analyzing and controlling instruments Service sectors (e.g. financial services, communication services, insurance, hotels, health services)

19% 16% 12% 22% 14% 17%

137

early in the project (Iansiti and MacCormack, 1997; Buganza et al., 2009). The project managers answered these questions for the project each one had managed. All items had possible responses ranging from 1 to 7 as is detailed in the Appendix. A formative indicator of flexible project planning was made up of questions asking for the degree that the process (Iansiti and MacCormack, 1997), the schedule (Eisenhardt and Tabrizi, 1995; Moorman and Miner, 1998) and the budget (Lenfle and Loch, 2010) for the project were flexible. The project managers answered these questions for the project each one had managed. A reflective variable for technological turbulence based on Jaworski and Kohli (1993) was made up of questions about the degree to which changes in technology provided opportunities in the industry, the degree to which technological breakthroughs facilitated new product ideas, the degree to which there were technological developments in the industry, and the degree to which technology was changing rapidly in the industry. The business managers answered these questions about conditions in the firm’s environment. Factor analysis confirmed that the items used loaded on one factor with a Cronbach’s alpha of 0.9. To measure technological innovativeness we used a scale based on Gatignon and Xuareb (1997). This included items for the degree of new technology use, radical new functionality and incorporation of new knowledge. Project managers answered these questions. Factor analysis confirmed that the items loaded on one factor with a Cronbach’s alpha of 0.8. While project managers provided information about the organization of each innovation project, business managers answered questions about the performance of the new product under study in each case. A large number of questions based on the recommendations of Griffin and Page (1993, 1996) and Moorman and Rust (1999) were included in the business manager survey (as well as the project manager survey for inter-rater reliability checking), and exploratory factor analysis was used to identify a variable for project performance. The variable was made up of four items (see Appendix) about how well the project goals for performance were met, how well the technical performance met specifications, how well quality guidelines were met and how well the new product satisfied customer needs. Summary statistics for the independent and dependent variables along with pairwise correlations are shown in Table 2. The correlation between flexible project planning and flexible project specifications is quite low, which justifies treating these two types of flexibility separately. Although the primary focus of this research was on technology turbulence, market turbulence was also included as a control variable (Santos-Vijande and A´lvarez-Gonza´lez, 2007). This variable, based on Jaworski and Kohli (1993), included questions about the degree to which customers’ product preferences were changing, the degree to which customers tended to look for new products, opportunities to attract new customers, changing customer needs, and the degree to which new customers were served. The business managers answered these questions.

Table 2 Summary statistics for dependent and independent variables and pairwise correlations.

1 2 3 4 5

Project performance Flexible project planning Flexible project specifications External technology turbulence Technological innovativeness p o0.01. p o0.05.

nnn nn

Mean

SD

Min

Max

1

2

3

4

5.97 4.40 5.07 3.92 4.90

0.70 1.17 1.34 1.08 1.24

3.00 2.00 1.50 1.00 1.80

7.00 7.00 7.00 6.60 7.00

 0.13 0.18nn  0.04 0.17nn

0.04 0.17nn 0.29nnn

0.02  0.04

0.10

138

M. Candi et al. / Technovation 33 (2013) 133–141

Project duration (Lakemond and Berggren, 2006) and project size (Eisenhardt and Tabrizi, 1995), both of which can be expected to be related with flexibility strategies, were also included as control variables. Project managers were asked about the size of the project team, which was used as an indicator of project size, and the length of time from the beginning of the project to new product launch in months. 3.3. Reliability, validity and post-estimation checks

Table 3 Regression results. Project Flexible performance project planning H1 & H2 H4 & H6 Flexible project planning  0.13nn Flexible product specifications 0.13nn Technology turbulence (external)  0.03 Technological innovativeness 0.17nnn (internal) Market turbulence  0.03 Project duration  0.01 Team size 0.10 Model significance (F) 2.88nnn 16% Proportion of variance explained (R2)

0.35nnn 0.27nn

Flexible project specifications H3 & H5

0.25nn  0.23nn

Reliability is the extent to which a measuring procedure is repeatable, i.e. yields the same or similar results in repeated tests. The rigorous process used to collect data from a random sample of Dutch firms is described in detail in Section 3.1. Thus, there can be reasonable confidence that if the same process were used to collect data from a different random sample of Dutch firms, repeated tests would yield similar results. The question of external validity, or generalizability to contexts outside of the Netherlands, was not addressed by this research and must be viewed as a limitation of the research. The research findings should be tested on similarly constructed samples from other countries. As discussed above, all the variables were measured by asking managers for their assessments. These are perceptions that must be acknowledged as having an element of subjectivity. However, existing research suggests that such subjective perceptions are reliable indicators of actual performance (e.g. Song and Parry 1997). One of the strengths of this research is that data were collected from two independent respondents for each innovation project. To enable testing inter-rater reliability, both project managers and business managers were asked to rate project performance. Although the business managers’ assessments were those used to generate the dependent variable for project performance, there was actually a high degree of inter-rater reliability between the two groups of managers. This was tested by generating the Kappa statistic for each of the four items making up the project performance variable. All were found to be statistically significant, suggesting that there was good inter-rater reliability between the two sets of managers. The Cronbach’s alpha statistic for each of the reflective variables was well over the generally accepted cut-off of 0.7 suggesting reliability in each of the variables. Furthermore, discriminant validity between the reflective variables was supported since an examination of correlations between the items used showed higher correlations between items belonging to each variable than items belonging to the other variables (Fornell, 1992). Regression post-estimation included checking variance inflation factors to verify that multicollinearity was not a problem. The maximum variance inflation factor was 1.4, which is well below the conservative threshold of 5 (Marquardt, 1970). The Breusch–Pagan/ Cook–Weisberg test was used to check for potential problems of heteroskedasticity. The test yielded small chi-squared values for all regressions, thus confirming that heteroskedasticity was not a problem. To test whether it was likely that the regression models were missing variables, Ramsey regression specification-error tests for omitted variables were conducted for each of the regressions and all suggested that additional variables were not needed.

customers are more likely to accept a product developed based on flexible specifications—so based on specifications that are allowed to change over the course of development—but less likely to accept a product developed using a flexible plan (process, schedule, and budget). Thus, Hypotheses 1 and 2 are supported. To investigate the conditions under which firms are more likely to adopt strategies of flexible project planning and flexible product specifications, regression analysis was conducted with external technological turbulence and project technological innovativeness as independent variables and flexible project planning and flexible project specifications, respectively, as dependent variables. The results are shown in Table 3. External technological turbulence was found to be positively related with the application of both flexible product specifications and flexible project planning, which lends support to Hypotheses 3 and 4. Project technological innovativeness was found to contribute negatively to flexible specifications and positively to flexible planning. So, Hypotheses 5 and 6 were supported by the data. The control variable for market turbulence was not found to contribute significantly to any of the models, which suggests that technology turbulence is a more important determinant of organization strategy. This resonates with the findings of Sethi and Iqbal (2008) and the model presented by Tatikonda and Montoya-Weiss (2001), in which market turbulence is posited to exert an influence after project completion. Project duration was not statistically significant in any of the models, which suggests that project performance and degree of flexibility in organization are not dependent on the length of a project. Team size was found to be statistically significant in one of three models. The larger the team size, the greater the degree of flexibility in project specifications. This suggests that the more people involved in an innovation project the more likely that the specifications will change over the course of the project. This seems intuitively correct since more people are likely to bring more perspectives and, thereby, more propensity for change.

4. Findings

5. Discussion

Ordinary least-squares regression was conducted to test the hypotheses. We first investigated the effects of flexible project planning and flexible product specifications on project performance and found statistically significant effects in the expected directions (Table 3). Flexible project planning contributes negatively to project performance, while flexible product specifications are positively related with project performance. This suggests that

5.1. Implications for theory

 0.08 0.07 0.08 3.05nn 12%

 0.16 0.10 0.54nnn 4.90nnn 17%

p o 0.01. po 0.05

nnn nn

Uncertainty is intrinsic and generally high in innovation projects (Cooper, 2011). Uncertainty is particularly high when there is a high level of turbulence within and outside the innovating organization. Existing research suggests that increased project flexibility is a common reaction to high levels of turbulence. However, existing

M. Candi et al. / Technovation 33 (2013) 133–141

research is characterized by definitions of project flexibility that are quite varied and sometimes unclear. In response to this shortcoming of existing work, our research makes an important distinction, namely between flexible project planning and flexible product specifications. Both types of flexibility have been included in existing definitions of flexibility, but as suggested by our research, they are conceptually different and make opposing contributions to performance. Our data suggest that the simultaneous adoption of flexible project planning and flexible product specifications is not common since the correlation between these variables is quite low (see Table 2). Thus, combining the two into broader variables for flexibility, which is commonly seen in much existing research may mask some of the effects and may also partially explain the contradictions apparent in existing research. Several researchers have found evidence suggesting that the flexibility to adapt product specifications to changing requirements is important for success (MacCormack et al., 2001; Iansiti, 1995; Iansiti and MacCormack, 1997; Thomke, 1997). Our findings that flexible product specifications are positively related with project performance resonate with this existing research. Flexible product specifications allow companies to adjust to changes in the environment, including, for example, new developments in technology or shifts in customer preferences. Conversely, flexible project planning was found to be negatively related with project performance. A major criticism of flexible project planning is the propensity to continue to wait for new information, which can result in severe delays (Pina e Cunha and Gomes, 2003). Furthermore, flexible project planning may reduce the urgency for innovation teams to work efficiently and effectively and keep the time spent on a project short and costs low. Therefore, while flexibility in project planning may facilitate the adaptation of project contents to evolving market requirements, it may not necessarily lead to improved performance outcomes. Indeed, our findings suggest that flexible project planning contributes negatively to project performance. This may be seen to contradict existing research that argues that a flexible approach to project planning is more effective than an inflexible approach, particularly under turbulent conditions (Eisenhardt and Tabrizi, 1995; Moorman and Miner, 1998; Li et al., 2010). However, the apparent contradiction may be explained by the fact that in this existing research flexibility is conceptualized both in terms of flexible project specifications and flexible project planning, while in our research we make an explicit distinction between the two types of project flexibility. Existing research has made much of the notion that flexibility is more prevalent or more effective under conditions of turbulence, but in most cases only one source of turbulence is considered. Turbulence can originate outside a firm or inside a firm. In this research we make an important explicit distinction between technological turbulence in a firm’s environment and the internal turbulence caused by radical innovations in technology. Our findings suggest that external technology turbulence is likely to lead to both types of flexibility, while internal technology turbulence is likely to lead to more flexible product planning and less flexible project specifications. The choice for flexible product specifications under conditions of external technology turbulence is consistent with existing research such as that of Moorman and Miner (1998). The finding that internal technology turbulence due to radical innovation has a negative relationship with flexible project specifications may appear to contradict existing research, such as Eisenhardt and Tabrizi (1995), but when we take into account that their research does not distinguish clearly between external and internal turbulence, nor between flexible project specifications and flexible project planning, the contradiction becomes less acute. To summarize, an important contribution of our research to the discourse on the effectiveness of flexibility in innovation lies in the distinction made between flexible project planning and flexible product specifications. Flexible product specifications can

139

facilitate the adaptation of a product to market conditions, and thus are likely to positively impact project performance, whereas flexible project planning does not support this process, while it may lead to delays and cost overruns, leading to higher prices, later-than-competitor market introduction, and thus to lower project performance. Also important is the distinction we make between turbulence originating outside a firm and turbulence originating inside a firm. In particular, we find that while external turbulence is likely to increase the propensity for both types of flexibility, internal turbulence is likely to lead to flexibility choices that may contribute negatively to project performance. 5.2. Implications for practice Our findings suggest that adopting a strategy of flexible product specifications is advisable, but that adopting a strategy of flexible project planning may not be as wise (cf. Olausson and Berggren, 2010). Interestingly, this resonates with the idea of agile development (Dyba˚ and Dingsøyr, 2008), which has enjoyed great popularity particularly in software development (Abrahamsson et al., 2003). Cockburn and Highsmith (2001) describe agile as a strategy that improves development efficiency to make teams more capable of responding to change in a timely fashion and maximizing their throughput. As suggested by research on agile software development, flexibility in product specifications (flexible project specifications) needs to be accompanied by micro-management, dayto-day work controls and continuous reporting on progress (inflexible project planning) to improve cost control and on¨ and Runeson, 2005), which implies rigid time delivery (Karlstrom project planning coupled with flexible project specifications. However, simply putting into place formal processes for innovation may not be sufficient since their existence does not necessarily insure that they will be adhered to (Christiansen and Varnes, 2009). Thus, deliberate project management is called for. It is important to insure balance between flexible project specifications and structured project planning (Olausson and Berggren, 2010). For example, in the context of computer-aided design (CAD) systems, Fixson and Marion (2012) found that even though digital design tools may help innovation processes because they allow for flexibility in specifications, these tools also can erode process discipline since they increase the propensity for continued and late changes, simply because they are possible. Cardinal et al. (2011) argue that project performance is not a function of interaction contingency between the environment and project organization as is commonly thought. Instead, the technological environment shapes performance priorities (e.g. meeting cost or time-to-market targets), which in turn influences project management, and ultimately performance outcomes (Cardinal et al., 2011). Our findings suggest that the technological environment does indeed shape a firm’s level and type of project flexibility, which in turn contributes to performance. More specifically, the results of our research suggest that firms need to avoid the pitfall of allowing external technology turbulence or internal technology turbulence to push them towards adopting flexible project planning, which could compromise the success of their project. At the same time, if dealing with a project with a high degree of innovativeness, they should resist the kind of single-mindedness in terms of product specifications that could likewise compromise the success of their project. We furthermore find that, unlike technology turbulence, market turbulence is related with neither flexible project specifications nor flexible project planning (for similar findings see Sethi and Iqbal (2008)). Thus, firms may limit themselves to managing the influences of external and internal technology turbulence during innovation. However, as suggested by existing literature, market turbulence may

140

M. Candi et al. / Technovation 33 (2013) 133–141

exert an influence after project completion (Tatikonda and MontoyaWeiss, 2001), suggesting that it may not be safe to ignore market turbulence throughout the product life cycle.

5.3. Limitations and directions for future research The empirical data on which this research is based have the strength of being derived from dual respondents. Project managers answered questions about the innovation process, while business managers answered questions about the success of the same innovation projects. The project manager questions were phrased in the past tense, meaning that the project managers were asked about an innovation project that had already been completed. Business manager questions were phrased so that they provided information about perceived project success at a time about 12–18 months after completion of the innovation projects. However, even given this implicit time lag, the data available made it possible to relate project organization only to short-term performance. Ideally, this kind of analysis should include data collected over a longer period of time so that both short-term and longer-term outcomes might be measured. Another potential limitation is that some important category of flexible organization may have been overlooked by the research. However, our extensive review of existing literature allowed us to group the types of flexibility included in this research quite naturally into the two types we propose. Nevertheless, the possibility of a missed dimension cannot be discounted entirely. Although we found a negative contribution of flexible project planning to project performance, there may be specific manifestations of flexible project planning that do work

positively. Future research could examine sub-categories of flexible project planning to provide a more complete picture of the forces at work. Finally, as our research was conducted only among Dutch firms, the question of external validity to other countries remains open. This article provides a detailed description of the process used to collect data from a random sample of innovation-active firms in a broad range of sectors as well as details about survey items and variables. Thus, the article should provide sufficient information for replication in other contexts.

Acknowledgments This research was funded in part by the Association of Dutch Designers, the Dutch Ministry of Economic Affairs and Pictoright.

Appendix A. Survey items and variables Table A1. References Abrahamsson, P., Juhani, W., Siponen, M.T., Ronkainen, J. 2003. New directions on agile methods: a comparative analysis. In ICSE ’03: Proceedings of the 25th International Conference on Software Engineering, , Washington DC, pp. 244–254. Cockburn, A., Highsmith, J., 2001. Agile software development: the people factor. Computer 34 (11), 131–133. Bhattacharya, S., Krishnan, V., Mahajan, V., 1998. Managing new product definition in high dynamic environments. Management Science 44 (11), 50–64. Bessant, J., Lamming, R., Noke, H., Phillips, W., 2005. Managing innovation beyond the steady state. Technovation 25 (12), 1366–1376.

Table A1 Variable

Items

References

Variable type

Flexible product specifications

To what extent were the requirements for [X] flexible? To what extent did management allow modifications to the project brief during the project? To what extent was the project concept frozen during the idea phase? To what extent was the project concept frozen during the development phase? To what extent was the process for developing [X] flexible? To what extent was the schedule for developing [X] flexible? To what extent was the budget for developing [X] flexible?

Iansiti and MacCormack (1997), Eisenhardt and Tabrizi (1995), Souder et al. (1998), Buganza et al. (2009), Lenfle and Loch, (2010)

Formative Eigenvalue ¼ 2.2, variance inflation factor ¼2.0

1–7: not at all—to Project a very great extent manager

Iansiti and MacCormack (1997), Eisenhardt and Tabrizi (1995), Moorman and Miner (1998), Lenfle and Loch (2010)

Formative Eigenvalue ¼ 1.5, variance inflation factor ¼1.2 Reflective Eigenvalue ¼ 2.0, Cronbach’s Alpha ¼0.8

1–7: not at all—to Project a very great extent manager

Flexible project planning

Technological To what extent is [X] a major improvement innovativeness compared with currently available products? To what extent did the development of [X] incorporate a large new body of knowledge? Overall, to what extent is [X] totally different from your main competitors’ products? To what extent does [X] incorporate technology not incorporated by competing products? Compared with competing products, to what extent does [X] provide radically different functionality? Technological Changes in technology provide big opportunities turbulence in our industry. A large number of new product ideas have been made possible through technological breakthroughs in our industry. There are major technological developments in our industry. The technology used in our industry is changing rapidly. Project The project’s goals for project performance were performance met completely. The technical performance of [X] meets the project specifications completely. [X] met quality guidelines completely. [X] satisfies customers’ needs.

Gatignon and Xuareb (1997)

Reliability

Response options

Respondent

1–7: not at all—to Project a very great extent manager

Jaworski and Kohli (1993)

Reflective Eigenvalue ¼ 3.5, Cronbach’s Alpha ¼0.9

1–7: disagree completely—agree completely

Business manager

Griffin and Page (1993, 1996), Moorman and Rust (1999)

Reflective Eigenvalue ¼ 1.7, Cronbach’s Alpha ¼0.8

1–7: disagree completely—agree completely

Business manager

M. Candi et al. / Technovation 33 (2013) 133–141

Bessant, J., von Stamm, B., Moeslein, K.M., Neyer, A.-K., 2010. Backing outsiders: selection strategies for discontinuous innovation. R&D Management 40 (4), 345–356. Blindenbach-Driessen, F., Van Dalen, J., Van den Ende, J., 2010. Subjective performance assessment of innovation projects. Journal of Product Innovation Management 27 (4), 572–592. ¨ ¨ Brattstrom, A., Lofsten, H., Richtne´r, A., 2012. Creativity, trust and systematic processes in product development. Research Policy 41, 743–755.Buganza, T., Dell’ Era, C., Verganti, R., 2009. Exploring the relationships between product development and environmental turbulence. The case of mobile TLC services. Journal of Product Innovation Management 26, 308–321. Buganza, T., Dell’Era, C., Verganti, R., 2009. Exploring the relationships between product development and environmental turbulence: the case of mobile TLC services. Journal of Product Innovation Management 26, 308–321. ¨ Calantone, R., Garcia, R., Droge, C., 2003. The effects of environmental turbulence on new product development strategy planning. Journal of Product Innovation Management 20 (2), 90–103. Cardinal, L.B., Turner, S.F., Fern, M.J., Burton, R.M., 2011. Organizing for product development across technological environments: performance trade-offs and priorities. Organization Science 22 (4), 1000–1025. Christiansen, J.K., Varnes, C.J., 2009. Formal rules in product development. Sensemaking of structured approaches. Journal of Product Innovation Management 26, 502–519. Cooper, R.G., 1994. Third-generation of new product processes. Journal of Product Innovation Management 11, 3–14. Cooper, R.G., 2009. How companies are reinventing their idea-to-launch methodologies. Research Technology Management, March April, 47–57. Cooper, R.G., 2011. Winning at New Products: Accelerating the Process from Idea to Launch, 4th edition Basic Books, New York. Diamantopoulos, A., Winklhofer, H.M., 2001. Index construction with formative indicators: an alternative to scale development. Journal of Marketing Research, 269–277. Dickson, P.R., 1997. Marketing Management, 2nd edition The Dryden Press, New York. Drejer, A., 2002. Towards a model for contingency of management of technology. Technovation 22 (6), 363–370. ˚ T., Dingsøyr, T., 2008. Empirical studies of agile software development: Dyba, a systematic review. Information and Software Technology 50, 833–859. Eisenhardt, K.M., Tabrizi, B.N., 1995. Accelerating adaptive processes. Product innovation in the global computer industry. Administrative Science Quarterly 40, 84–110, March. Ferna´ndez, P., Del Rio, L., Varela, J., Bande, B., 2010. Relationships among functional units and new product performance: the moderating effect of technological turbulence. Technovation 30 (6), 310–321. Fixson, S.K., Marion, T.J., 2012. Backloading: a potential side effect of employing digital design tools in new product development. Journal of Product Innovation Management 29 (S1), 140–156. Fornell, C., 1992. A national customer satisfaction barometer: the Swedish experience. Journal of Marketing 56, 6–21. Garud, R., Jain, S., Tuertscher, Ph., 2008. Incomplete by design and designing for incompleteness. Organization Studies 29 (03), 351–371. Gatignon, H., Xuareb, J.M., 1997. Strategic orientation of the firm new product performance. Journal of Marketing Research 34 (1), 77. Griffin, A., Page, A.L., 1993. An interim report on measuring product development success and failure. Journal of Product Innovation Management 10, 291–308. Griffin, A., Page, A.L., 1996. PDMA success measurement project: recommended measures for product development success and failure. Journal of Product Innovation Management 13, 478–496. Iansiti, M., 1995. Technology integration: managing technological evolution in a complex environment. Research Policy 24 (4), 521–542. Iansiti, M., MacCormack, A., 1997. Developing products on internet time. Harvard Business Review (September–October), 107–117. Jaworski, B.J., Kohli, A.K., 1993. Market orientation: antecedents and consequences. Journal of Marketing 57 (3), 53.

141

¨ Karlstrom, D., Runeson, P., 2005. Combining agile methods with stage-gate project management. IEEE Software 22 (3), 43–49. Kawasaki, G., 1999. Rules for Revolutionaries. Collins Business, New York. Knot, J.M.C., van den Ende, J.C.M., Vergragt, P.J., 2001. Flexibility strategies for sustainable technology development. Technovation 21, 335–343. Krishnan, V., Eppinger, S.D., Whitney, D.E., 1997. A model based framework to overlap product development activities. Management Science 43 (4), 437–451. Lakemond, N., Berggren, C., 2006. Co-locating NPD? The need for combining project focus and organizational integration. Technovation 26 (7), 807–819. Langerak, F., Hultink, E.J., 2006. The impact of product innovativeness on the link between development speed and new product profitability. Journal of Product Innovation Management 23, 203–214. Lee, K.B., Wong, V., 2011. Identifying the moderating influences of external environments on new product development process. Technovation 31 (10–11), 598–612. Lenfle, S., Loch, C., 2010. Lost roots: how project management came to emphasize control over flexibility and novelty. California Management Review 53 (1), 32–55. Levardy, V., Browning, T.R., 2009. An adaptive process model to support product development project management. IEEE Transactions on Engineering Management 56 (4), 600–619. Li, Y., Su, Z., Liu, Y., 2010. Can strategic flexibility help firms profit from product innovation? Technovation 30 (5), 300–309. 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. Marquardt, D.W., 1970. Generalized inverses, ridge regression, biased linear estimation, and nonlinear estimation, Technometrics 12, 591–612. Moorman, C., Miner, A.S., 1998. The Convergence of planning and execution: improvisation in new product development. Journal of Marketing 62, 1–20. Moorman, C., Rust, R.T., 1999. The role of marketing. Journal of Marketing 63, 180–197. Olausson, D., Berggren, C., 2010. Managing uncertain, complex product development in high-tech firms: in search of controlled flexibility. R&D Management 40 (4), 383–399. Pina e Cunha, M., Gomes, J.F.S., 2003. Order and disorder in product innovation models. Creativity and Innovation Management 12 (3), 174–187. Santos-Vijande, M.L., A´lvarez-Gonza´lez, L.I., 2007. Innovativeness and organizational innovation in total quality oriented firms: the moderating role of market turbulence. Technovation 27 (9), 514–532. Schlesinger, L.A., Kiefer, C.F., Brown, P.B., 2012. New project? Don’t analyze—act. Harvard Business Review 2012, 154–158, March. Sethi, R., Iqbal, Z., 2008. Stage-gate controls, learning failure, and adverse effect on novel new products. Journal of Marketing 72, 118–134, January. Song, X.M., Parry, M.E., 1997. The determinants of Japanese new product successes. Journal of Marketing Research 34, 64–76. Souder, W.E., Sherman, J.D., Davies-Cooper, R., 1998. Environmental uncertainty, organizational integration, and new product development effectiveness: a test of contingency theory. Journal of Product Innovation Management 15, 520–533. Tatikonda, M.V., Montoya-Weiss, M.M., 2001. Integrating operations and marketing perspectives of product innovation. The influence of organizational process factors and capabilities on development performance. Management Science 47 (1), 151–172. Thomke, S., Reinertsen, D., 1998. Agile product development: managing development flexibility in uncertain environments. California Management Review 41 (1), 8–30. Thomke, S.H., 1997. The role of flexibility in the development of new products: an empirical study. Research Policy 26, 105–119. Veryzer, R., 1998. Discontinuous innovation and new product development process. Journal of Product Innovation Management 15, 304–321. Wheelwright, S.C., Clark, K.B., 1992. Revolutionizing product development. The Free Press, New York.