Action Planning for New Product Development Projects

8/7/2019 Action Planning for New Product Development Projects

1/15

Action Planning for New Product

Development ProjectsJan Buijs

Two empirical studies are presented to show how experienced project leaders execute NewProduct Development (NPD) projects. In the first study we interviewed project leaders fromfour different design firms. We discovered that inside realistic NPD projects the NPD activi-ties seldom occur in the same order as they are described in the NPD literature. Some activitiesare omitted, some activities are run in parallel and some even have a seemingly illogicaltiming. The reasons for these strange patterns are usually project-specific. The NPD project

leaders distinguish four types of NPD projects. On the one hand, familiar (client well knownand/or standard technology and/or re-design) or non-familiar projects (new client and/or newtechnology and/or innovative design), and on the other hand, the complexity of the product(simple versus complex), and they plan their NPD projects differently according to those fourtypes. For instance, within simple and familiar projects they omit more NPD activities than inprojects with a more complex and new nature. In the second empirical study we did a matchedpairing study (finding NPD projects which would match each of the four types). This time weinterviewed experienced project leaders from different companies, because they are probablymore familiar with only one type of NPD project. We found a minimal and a regular NPDprocess. Projects on new products (the non-familiar type) contain the most activities in thetotal project. Complex projects execute more activities in the first stages, and also differentactivities than in non-complex projects. We also found that NPD project leaders adapt anopportunistic attitude towards carrying out activities in parallel in order to gain time.

Introduction

The New Product Development (NPD)process or Product Creation Process, ormore generally the Product InnovationProcess, has been subject to model builderssince the early to mid 1950s. The complexprocess of opportunity finding, idea genera-tion, concept building, prototype testing andthe market introduction of a new product has

been described in many ways. However, anewly appointed project leader gets little orno support from the theory on how to plan hisor her specific NPD project.

Some 24 years ago, Saren published an over-view of the different types of models prescrib-ing the different steps or activities within theproduct innovation process (Saren, 1984). Hedistinguished five different categories:

1. Departmental stage models;2. Activity-based stage models;3. Decision-based models;

4. Transformation models;5. Stimulus response models.

He did not qualify the one category to be betterthan any other category. He also did notprovide any help in selecting a category for aspecific NPD project. It is all up to the projectleader to find his or her way through this jungle of different models.

Since Sarens paper, no new categories have

been developed (for examples of the models,see for instance the Germans Pahl & Beitz(1984), the Danes Andreasen & Hein (1985),the Englishman Cross (1994), the DutchmenRoozenburg & Eekels (1995) and Buijs & Valk-enburg (2005) or Coopers Stage-Gate Model(1984) from Canada and Ulrich & Eppinger(1995) from the United States). Most of thesemodels come down to a variation of a stage-gate process. The total NPD process is dividedinto a series of stages (e.g., pre-development,development and market introduction), andinside each stage a couple of interrelated

ACTION PLANNING FOR NPD PROJECTS 31

Volume 17 Number 4 200doi:10.1111/j.1467-8691.2008.0049

2008 The AuthorJournal compilation 2008 Blackwell Publishing


2/15

actions have to be executed. Each stage iscompleted if a pre-determined milestone isreached (e.g., market feasibility or prototypetesting). This milestone is called a gate. Aftergetting permission (usually by higher manage-ment) to go through the gate, the next stagewill be executed. If permission is not granted,usually an iteration takes place; in other wordsan earlier stage or stages is executed once moreuntil the necessary quality to get through thegate has been reached. This continues until thenew product has entered the market. Some-times the results of a stage lead to the end ofthe NPD project.

The academic debate on these models has been concentrated on the one hand on thenumber of stages, the number and quality ofthe gates, which activities belong to whichstage, and on the other hand on where theinnovation process starts (with a technologicalidea or with the recognition of a need in the

market place, a so-called opportunity) andwhere it ends (the first product sold or withthe satisfaction of a happy customer or con-sumer?). Now the most recent models haveadded the fuzzy front end at the beginning ofthe process, and some have added the use ofthe product at the end. The most advancedones have even tied together the product usestage and the fuzzy front end and come upwith a circular innovation model (for an over-view of this development, see Buijs, 2003). Inconclusion, the theoretical innovation processhas been made longer, both up front as wellas downstream.

All prescriptive models use their own jargon, use their own graphic representationsand fail to show all the necessary iterations thereal innovation process is famous for. All showmore or less rational and logical sequencesof actions. In reality, however, we rarely eversee these rational step-by-step sequences asshown in the theory books.

As long ago as 1983, Cooper published aninvestigation about the empiricism of NPD,and discovered that seven different patternscan occur:

1. The market oriented process.2. The design oriented process.3. The balanced complete process.4. The front end dominated process.5. The minimum process.6. The launch with prototype process.7. The prototype dominated process.

His research was based on an inquiry amonginnovation managers. He gave them a list of20 different NPD activities known from thetheory books, and asked whether an activityhad taken place in their NPD process, in whichsequence and what the duration time was of

the executed activities. He could not give anyexplanation about the reasons behind thesepatterns.

Comparable results were found in a studyin the Netherlands among about 150 innovat-ing small and medium-size enterprises (SMEs)(Buijs, 1984, 1987). In this study, eight differentpatterns were observed, not based on a post-facto analysis as in the Cooper study, but basedon real-time observations. But once again,here also there was no explanation for thesedifferences.

Now, nearly 25 years later, we felt the needto once again dive deep into the real world ofNPD projects, and to redo some of thesestudies and now hopefully find some cluesabout the reasons behind these differentpatterns.

We carried out two empirical studies, onewith experienced project leaders from fourdesign agencies, and a second one also with

experienced project leaders but now from11 different companies all with their ownin-house design/development department.Based on the interviews with NPD projectleaders from the design firms, we discoverednot only that experienced project leaders rec-ognize different NPD patterns, but that theyeven have reasons to distinguish NPD projectsaccording to two different aspects. The firstaspect is the complexity of the design task, andthe second aspect is the familiarity of thedesign task. This leads to four different catego-ries of NPD projects: simple and familiar,complex and familiar, simple and new andfinally complex and new.

In our second study, we searched for caseswhich matched these four categories. We inter-viewed experienced project leaders in 11 dif-ferent firms about their NPD projects. Weselected only projects which resulted in designaward-winning products (to be certain of theperformance of the NPD projects investi-gated). We asked the project leaders about thenumber of activities, what activities wereexecuted in parallel and which had to besequential, and how much time the activitiestook. We did indeed find different patterns

of NPD processes for the four different catego-ries of projects.

Preliminary Study: NPD Activitiesfrom the Literature

In an analysis of the differences in modellingproduct innovation processes, a comparison between 90 different innovation models pub-lished in the NPD literature was made byVan der Zee (2003). The models were from allover the world and from the early 1950s to

320 CREATIVITY AND INNOVATION MANAGEMENT

Volume 17 Number 4 2008 2008 The Author

Journal compilation 2008 Blackwell Publishing


3/15

the present day. We discovered that, in total,more than 1,248 different terms were used todescribe specific product innovation activities.We condensed these 1,248 into a set of 54innovation activities.

This proved to be a difficult task becausenone of the original authors is very explicitabout the meaning of the terms used. Forinstance, what is a product idea? Just wordson a piece of paper, a design sketch, or a fullydocumented idea for a new business activityincluding concrete market and technologicalinformation. Or take the notion of a proto-type. For engineers and product designers, aprototype is a physical, tangible object you canperform tests with. According to MichaelSchrage in his book Serious Play (2000), evenExcel spreadsheets function as a prototype.Or, is the meaning of the German word Grund-analyse exactly the same as Feasibility studyor has it more to do with Scanning the com-

petitive environment?The way we did this convergence was to

map all models against each other. The walls ofour study room was covered with photocopiesof all the models. All the models have atleast two or three different terms for the sameactivity in real-life NPD. For instance, productidea and ideation, detailed design andembodiment design, market introductionand product launch. We positioned all themodels against each other, using these termswith more or less equal meanings as thelinking pins.

Some models start with a product idea,some with the corporate strategy; so the start-ing points of the models can differ greatly.Other models have different endings: somestop at the prototype, some include productuse, and others include maintaining and recy-cling of the product. Recent models includethe so-called fuzzy front end of innovation(Koen et al., 2001), the older models restrictthemselves to a more limited engineering viewon innovation (= product development). Anddepending on this length of the modelledinnovation process, the models are detailed toa greater or lesser degree. Long models tend

to be more abstract, with fewer details; shortmodels tend to be more concrete with a lot ofdetails. Some models distinguish only sevenstages (VDI 2221, 1986), others nearly 50(Archer, 1971).

As already stated, the convergence from1,248 different terms to 54 more or less sharedterms was a difficult one, but nonetheless avery interesting process with many discus-sions. We used triangulation to come up withthis list. Three researchers did the convergenceindependent of each other. Then the resultswere compared. When there were differences

a discussion was held. This discussion wasguided by a fourth researcher. At the end,all four researchers accepted the result. Thecomplete list is shown in Table 1.

The sequence of the 54 activities in our list isbased on all the sequences seen in all the dif-ferent models, and remains open for debate.For instance, after appointing a project leader(which is activity no. 5 on our list), you caneasily imagine that the formation of the projectteam will be the next activity. In our con-densed list forming the team is activity no.11. This could imply that the project leader willexecute all the in-between activities (nos 610)individually. But it could well be that thoseactivities are carried out by the responsiblefunctional departments of the company andthat their results will be communicated to theteam that is formed later. After its formationthe innovation team can build on these results.

The total list of 54 NPD activities can be

divided into three groups: the pre-, core- andpost-NPD activities. We noticed four activitiesat the very beginning of the fuzzy front end(or rather activities which should be executedeven before starting a NPD project, such asdoing basic technological research, hiring newstaff or scanning the competitive environ-ment). These activities are labelled the pre-NPD activities, our numbers 14. We see 42activities as the core NPD activities (nos 546),and eight activities which have to do with theimplementation of the results of the NPDproject (activities such as manufacturing, dis-tributing, maintenance or recycling, nos 4754).These are labelled the post-NPD activities.

First Empirical Study: ExploringCategorizations of NPD Projects

This convergence from theoretical models into54 NPD activities does not say anything aboutduration time per activity or step, does not sayanything about parallel processing of differentactivities in certain stages of the NPD process,nor does it say anything about the possibilityof skipping certain activities. Based on the

NPD theory, no activities can be skipped, butin real-life NPD we see that in some casesactivities are deliberately skipped, accidentallyneglected or stupidly just forgotten. And stillnew products enter the market after suchincomplete NPD processes. Some of themeven turn into major market successes.

Cooper already indicated in his researchthat different patterns of innovation processesdo exist in real-life NPD (Cooper, 1983). Someof the innovation activities are carried out inparallel, others sequentially. And not alwaysall theoretical NPD activities are executed.





4/15

Table 1. The Total List of Acknowledged NPD Activities

1. Development of companys competences by carrying out technological research, hiring newstaff, etc.

2. Stimulating the generation of new ideas, or new applications of existing ideas, etc.3. Carrying out explorative market research by scanning the competitive environment, looking for

trends in the market, searching for needs, scientific developments, etc.4. Adjusting the strategic plans of the company5. Appointing a project leader6. Preliminary description of the idea for a new product or a new challenge in the marketplace7. Analysing present product portfolio of the company8. Development of preliminary programme of requirements of the new product9. Investigating the commercial feasibility of the new product

10. Generating principal solutions for the new product11. Forming the project team12. Determination and analysis of the target group13. Building and testing of experimental prototypes14. Generating and evaluating ideas for the new product15. Preliminary planning for the total NPD process16. Patent search17. Carrying out the technical feasibility of the new product18. Developing and testing concepts for the new product19. Analysing competitive products20. Developing the manufacturing plan for the new product (make-or-buy decision)21. Developing the promotion plan for the new product22. Detailed design of the new product23. Carrying out user tests24. Finding and selecting suppliers25. Building prototype26. Testing prototype27. Building manufacturing facilities28. Making technical drawings of the new product29. Developing tooling30. Testing new manufacturing facilities31. Debugging manufacturing process32. Sales forecasting of the new product33. Debugging the new product34. Coaching the first production runs35. Developing maintenance plans for the new product36. Training staff in assembling the new product37. Testing promotion plan for the new product38. Developing maintenance documents for the new product39. Training staff in installing the product40. Promoting the new product41. Training staff in maintaining the new product42. Developing the distribution plan for the new product43. Training users44. Evaluating the NPD process45. Certifying the new product46. Patenting the new technology47. Manufacturing the new product48. Assembling the new product49. Distributing the new product50. Selling the new product51. Installing the new product52. Maintenance and servicing the new product53. Recycling the new product54. Disposal of the new product





5/15

According to Coopers empirical research, theminimal NPD process will have at least twodifferent sequential activities (product designand development and market launch).Coopers maximum NPD process scoredonly nine different activities. Keep in mindthat the abstraction levels on which the differ-ent NPD activities are described determine thelengths of the different steps. Coopersproduct design and development covers atleast activities 8, 9, 10, 13, 14 and 17 in our list.And you can even add activities 17, 22, 25 and26 if you want to. So what does a minimalprocess of only two steps mean? Comparingthe different NPD models will remain tricky.

In our first empirical study we interviewedfour project leaders of NPD projects. Allproject leaders are experienced productdesigners (more than five years of professionalpractice) and are working at four differentDutch design agencies. They were chosen for

interview because they had worked on manydifferent NPD projects for a variety of clients.We asked them to look at the activities theyhad executed during their NPD projects, andto compare those against the list of the 54theoretical activities.

We had made separate cards for each the 54activities and asked the project leaders if they(or their team) had performed this activity.Some extra cards were still blank, to allowthe project leaders to add other NPD activitiesto the list of 54.

All project leaders are working for differentdesign firms. For competitive reasons, thesefirms use their own specific words and modelsto describe their own NPD process. So wehad the same semantic problems with them aswe had encountered before during our ownconvergence from the NPD literature. Butbecause we now had experience with the dif-ferent terminologies, we could easily discusswith the project leaders the specific content ofthe NPD activities and were not restricted tothe labels that the design firms use to describetheir own way of executing a product innova-tion project for a client. In the end, no newNPD activities were discovered, and most of

the project leaders accepted the terminology inthe list of 54 NPD activities we had condensedfrom the literature.

The research methodology was to ask theproject leaders to place the NPD activities ona grid scorecard to find out the sequence ofactivities, the parallel execution of activitiesand more or less automatically they mentionedthe duration time per activity. After the projectleaders had placed all the relevant NPD activi-ties on the grid, the yellow stickers were gluedin place and the duration time was writtendown.

We borrowed the idea from Cooper not touse the real duration time of the NPD projectsto form patterns, but to use 100 per cent as theduration time and to place the activities withinthis 100 per cent frame. The reason for doingthis is that otherwise the patterns will be cat-egorized according to real duration time. Sothe shortest projects will all be categorizedtogether, and similarly with all the projectswith the longest duration time. The categoriza-tion would then be based on differences in totalduration time and not on differences in pat-terns of NPD activities, which is what we arereally interested in. In the past we have usedthis research method successfully (Buijs, 1984).

In order to do this analysis properly, wedeveloped a new version of the grid scorecard,adding five columns, each divided into 10segments of 10 per cent of the duration time ofthe total NPD process. The columns are usedto place more or less similar NPD activities

together. The five columns are from left toright:

1. Technology2. Product3. Market4. Organization5. Miscellaneous.

The grid scorecard is shown in Figure 1.This methodology is based on a specific

model of the innovation process in which ineach stage three groups of parallel NPD activi-ties are distinguished (Buijs, 2003). Central arethe product-related activities, such as develop-ing the programme of requirements, or generatingideas. To the left, which in this circular modelmeans the inside, are all technology-relatedNPD activities, such as manufacturing or build-ing prototypes. To the right, on the outside ofthe model, are all market-oriented NPD activi-ties, such as doing market research or making apromotion plan. It proved impossible to assignall 54 NPD activities to these three categories.For instance, appointing the project leader orforming the innovation team are of a differ-ent nature. Therefore, we added two extracolumns: one for organizational NPD aspects

and one category for the rest (= Miscella-neous). Now all 54 activities could be assignedto a relevant column. In this way, we couldanalyse the patterns of all different cases in anequal and comparable manner.

To show the results of this kind of analysiswe will show the patterns of two differentNPD projects, one for Product A and one forProduct B, each from a different design agency(see Figure 2).

The NPD process of Product A has the fol-lowing stream of activities. The reading startsfrom the top of the left figure (the start of the





6/15

Grid scorecard

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Duration

time

Technology Product Market Organization Miscellaneous

Figure 1. The Grid Scorecard with the Five Columns

10

8

8

663

3

5 57

11 119

9

10

1022

13

17 17

2526

2526

24

2820

29

27

27

30

31

30

31

48

38

33

34

3943

424443

3323

37 4546

19

2024

29 28

22

18 19

12

12

13

1515

14

20

30

40

50

60

70

80

90

100Technology Product Market OrganizationMiscellaneous Technology Product Market Organization Miscellaneous

%

10

20

30

40

50

60

70

80

90

100

%Product A Product B

Figure 2. Two Different NPD Patterns: Product A by Design Agency X (Left), and Product B by DesignAgency Y (Right)Notes: The numbers correspond with the numbered NPD activities of Table 1. An interestingphenomenon did occur while scoring the different NPD activities on the scorecards. To put thenumbers on the grid card, we used standard sheets with stick-on labels with numbers printed on.After sticking on the different NPD patterns on the grids, the unused stickers on the label sheetsshowed a kind of contrasting view in which the numbers left over more or less mirror the NPDpatterns.





7/15

NPD process) and from left to right over thefive columns. It ends in the bottom right-handcorner. In brackets we will add the type ofcolumn the activity is assigned to.

3. Carrying out explorative market researchby scanning the competitive environment,looking for trends in the market, searching

for needs, scientific developments, etc. [M]8. Development of preliminary programmeof requirements of the new product [P]

6. Preliminary description of the idea for anew product or a new challenge in themarketplace [P]

15. Preliminary planning for the total NPDprocess [P]

5. Appointing a project leader [O]19. Analysing competitive products [M]

9. Investigating the commercial feasibility ofthe new product [M]

11. Forming the project team [O]14. Generating and evaluating ideas for the

new product [P]12. Determining and analysis of the target

group [M]10. Generating principle solutions for the new

product [P]13. Building and testing of experimental pro-

totypes [T]17. Carrying out the technical feasibility of the

new product [T]22. Detailed design of the new product [P]25. Building prototype [T]26. Testing prototype [T]23. Carrying out user tests [M]

33. Debugging the new product [P]24. Finding and selecting suppliers [T]28. Making technical drawings of the new

product [T]37. Testing promotion plan for the new

product [M]20. Developing the manufacturing plan for

the new product (make-or-buy decision)[T]

45. Certifying the new product [Mi]46. Patenting the new technology [Mi]29. Developing tooling [T]27. Building manufacturing facilities [T]30. Testing new manufacturing facilities [T]

31. Debugging manufacturing process [T]48. Assembling the new product [T]43. Training users [M]

The NPD process for Product B (the right partof Figure 2) reads as follows:

6. Preliminary description of the idea for anew product or a new challenge in themarketplace [P]

15. Preliminary planning for the total NPDprocess [P]

7. Analysing present product portfolio of thecompany [Mi]

5. Appointing a project leader [O]11. Forming the project team [O]

8. Development of preliminary programmeof requirements of the new product [P]

3. Carrying out explorative market researchby scanning the competitive environment,looking for trends in the market, searchingfor needs, scientific developments, etc [M]

12. Determining and analysis of the targetgroup [M]

10. Generating principle solutions for the newproduct [P]

17. Carrying out the technical feasibility of thenew product [T]

19. Analysing competitive products [M]18. Developing and testing concepts for the

new product [P]13. Building and testing of experimental

prototypes [T]9. Investigating the commercial feasibility of

the new product [M]

20. Developing the manufacturing plan forthe new product (make-or-buy decision)[T]

22. Detailed design of the new product [P]24. Finding and selecting suppliers [T]29. Developing tooling [T]28. Making technical drawings of the new

product [T]25. Building prototype [T]26. Testing prototype [T]27. Building manufacturing facilities [T]33. Debugging the new product [P]30. Testing new manufacturing facilities [T]31. Debugging manufacturing process [T]34. Coaching the first production runs [O]38. Developing maintenance documents for

the new product [T]43. Training users [M]39. Training staff in installing the product [O]44. Evaluating the NPD process [P]42. Developing the distribution plan for the

new product [M]

Both processes show more or less normalsteps in the NPD process (i.e., as publishedin the NPD literature), more or less normalsequences, such as building manufacturing

facilities (no. 27), testing manufacturing facilities(no. 30) and debugging manufacturing process(no. 31). But both also show that certain activi-ties are run in parallel, such as carrying outexplorative market research by scanning the competitive environment, looking for trends in themarket, searching for needs, scientific develop-ments (no. 3), the development of preliminary pro-gramme of requirements of the new product (no.8), the preliminary description of the idea for a newproduct or a new challenge in the marketplace (no.6) and the preliminary planning for the total NPDprocess (no. 15). Both projects have used in total





8/15

30 different NPD activities to get the newproduct on the market. But those 30 are notthe same ones!

Now we zoom out to all four cases. If welook at the 42 core NPD activities, 41 of themwere executed at least once in those four cases.The only theoretical activity that was notexecuted was activity no. 35 (developing main-tenance plans for the new product). Only two outof the four pre-NPD activities were recorded,and only one out of the eight of the post-NPDactivities was recognized (no. 48 assembling thenew product).

We discussed our results with the projectleaders in individual feedback sessions andtalked about the possible explications for thedifferent patterns in the NPD projects. Fortheir own projects they had very specific argu-ments (for each case/project more or lessunique) why certain activities were omitted,run in parallel or had a special sequence. There

were no general arguments used by all four tofind reasons for specific patterns. But interest-ingly enough, when we asked about planningfuture NPD projects, all four used the sametype of reasoning. They all had two argumentsto plan a new project. One reason is the famil-iarity of the NPD project. Is it a redesignproject, is it a well-known client they hadworked for before or is the technology wellknown. The other reason is the complexity ofthe product. Is it a simple product, with veryfew components and with only one technol-ogy, or is it a complex product, with lots ofdifferent components, with a multitude ofsuppliers involved and with different ornew technologies?

Based on their work, these experiencedproject leaders distinguish four different cat-egories of NPD projects (see Figure 3). Theexperienced project leaders indicated that,

according to their judgement of the category ofa future NPD project, they adjusted their plan-ning of the activities of the project. Going fromtype A to D they suggested that the number ofactivities is increasing, and that the number ofparallel activities is decreasing. According totheir insights, the total duration time is alsoincreasing from type A to D. Planning NPDactivities for the next project is dependenton the type of project in this familiarity/complexity grid. And a simple product meansa simple project plan; a complex productneeds a complex project plan. We decided tostart another study to test this reasoning.

Second Empirical Study:Investigating Patterns inNPD Projects

Now, we know about the differences between

the theoretical models of the NPD process, andwe know about empirical evidence that realNPD projects are not executed according tothese models. But what we do not know arethe reasons behind these differences. Based onthe interviews with the experienced projectleaders, we have a hypothesis that for the fourdifferent categories of NPD projects differentpatterns of NPD activities might exist. Simpleproducts have simple project structures andcomplex products have complex structures.

Because these differences could be causedby differences in companies, markets or tech-nologies, we decided to look for cases whichare more or less equal in design quality. Inresearch methodological terms, we are lookingfor matched pairings (the empirical caseshould fit the theoretical category).

We have no real hypothesis about thecontent of the different patterns (simple

Type B:

Simple product & New projectType D:

Complex product & New project

The familiarity

of the

NPD projectType A:

Simple product & Familiar

project

Type C:

Complex product & Familiar

project

The complexity of the NPD product

Figure 3. The Four Categories of NPD Projects





9/15

product implies simple project and complexproduct implies complex project, but there isno theoretical base for it). So we decided to doan exploratory study to test the four categoriesof NPD projects and whether these categoriesnot only have different patterns, but also that,where these differences occur, they confirmthe familiarity/complexity grid.

In this second exploratory study, we inter-viewed 11 experienced NPD project leaders.This time we choose to interview projectleaders inside companies. The reason is thatproject leaders of design agencies might haveexperiences with all four categories of NPDprojects, which could confuse their memories.In this study we wanted to focus on differ-ences between the project categories. We alsohoped to find more of the pre- and post-NPDactivities because a company is responsible forthe whole NPD project whereas a design firmmight focus on only a part (as we have seen in

the exploratory study). All 11 project leaderscome from 11 different companies.

We were looking for projects in each of thefour categories of NPD projects. We wanted,for comparison reasons, to guarantee a certaindegree of design quality of each project andthe resulting new product. Therefore, wechoose to investigate NPD projects that had just won a Dutch design award for theirproduct. All our 11 cases are prize-winningproducts, although in different industrysectors. With this careful selection of the cases,we tried to have NPD projects under investi-gation which are of the same high designquality and therefore comparable.

The distribution of the cases over the fourcategories of NPD projects introduced yetanother difficulty. The four types are based ondifferences on two axes: one on familiarity ofthe project and the other on the complexity ofthe product (remember Figure 2). The com-plexity of the NPD product is observable fromthe outside; you can see it by looking at theproducts that are the result of the NPD project.We did a small test to see whether design stu-dents were able to judge the complexity ofproducts just by looking at them. For this test

we used the products from our first study (atthe design agencies), and we compared thestudents judgements with those of the origi-nal project leaders who designed the products.All the students had the same judgement asthe professionals.

We did the same experiment with thedegree of novelty of the project, and no studentwas able the make the right judgement. Thereason is quite simple: novelty is based on theknowledge and experience of the projectleader who is going to design the new product.And outsiders are not able to judge this aspect.

So in the case selection we had torely on theinformation from the project leaders them-selves to judge the scoring on the familiarityaxis in order to categorize the case into therelevant category. For the complexity axis wemade our own external judgement and com-pared that with those of the project leaders.As expected, there was a 100 per cent match.

To get cases, we categorized the prize-winning products on complexity, phoned thecompanies who designed these products, andasked them if they were willing to join ourresearch. After a positive reaction, we held apreliminary interview to find out about thefamiliarity scoring. After a couple of attempts,we finally had our cases matching the catego-ries. For reasons of confidentiality, we arenot allowed to show pictures of the productsand/or to give the names of the companies.

The research methodology used for gettingthe NPD activity patterns is the same as in our

first exploratory study. We asked the projectleaders about the number of activities, whichactivities were executed in parallel and whichwere carried out sequentially, and what theduration time of each different activity hadbeen.

We offered the interviewees our cards of the54 NPD activities, asked them to locate theactivities on the grid and to mention the dura-tion time per activity. We formatted everythingaccording to Coopers 100 per cent format andanalysed the results. The interviews wereexecuted at the offices of the project leaders.They had collected relevant data from theirproject files on starting dates of activities,number of participants, they showed ussketches and drawings and showed us the fin-ished end result of their NPD work: the newprize-winning product. They proved to bevery co-operative and open and were veryinterested in our study. The interviews usuallytook between 1.5 and 3 hours.

The Data

Our raw data for each case is a large piece of

paper with the 100 per cent grid, the yellowstickers with the (numbered) NPD activityglued in their positions. The duration time peractivity was also written on the stickers. Wealso had some background information aboutthe product involved, the composition of thedesign team and some information aboutthe company.

The 11 cases were evenly spread over thefour categories (see Figure 4). Cases A and Bexecuted 31 different NPD activities each. TheType B cases show a broader range: case C 29,case D only 22 (the least of all cases) and case E





10/15

33. The Type C cases are more or less equal:cases F and G 28 and case H 30. The Type Dcases are also diverse: case I 25, case J 34 (thecase with the most NPD activities) and case K

29. The two cases from our first study (seeFigure 2) scored 30 NPD activities each. Theaverage number of NPD activities for all 11cases is 29. Type A cases are above average,Type B cases are below average, Type C casesare average and Type D cases are just aboveaverage. An interesting finding is that in thecategory of the easier kind (Type A: simple andfamiliar), more NPD activities are executedthan in the more complex ones, which contra-dicts the preliminary thoughts of the projectleaders of the design agencies.

Results: the Existence ofNPD Activities

If we compare the 54 NPD activities from NPDtheory and the NPD activities executed in thevarious cases, we see the following results (thescore per activity is put in brackets in boldtype after the description of the activity):

1. Development of companys competences by carrying out technological research,hiring new staff, etc. (0)

2. Stimulating the generation of new ideas,or new applications of existing ideas, etc.

(1)3. Carrying out explorative market researchby scanning the competitive environment,looking for trends in the market, searchingfor needs, scientific developments, etc. (9)

4. Adjusting the strategic plans of thecompany (0)

5. Appointing a project leader (10)6. Preliminary description of the idea for a

new product or a new challenge in themarketplace (8)

7. Analysing present product portfolio of thecompany (6)

8. Development of preliminary programmeof requirements of the new product (11)

9. Investigating the commercial feasibility ofthe new product (9)

10. Generating principle solutions for the newproduct (5)

11. Forming the project team (11)12. Determining and analysis of the target

group (9)13. Building and testing of experimental

prototypes (9)14. Generating and evaluating ideas for the

new product (9)15. Preliminary planning for the total NPD

process (11)16. Patent search (7)17. Carrying out the technical feasibility of the

new product (11)18. Developing and testing concepts for the

new product (11)19. Analysing competitive products (9)20. Developing the manufacturing plan for

the new product (make-or-buy decision)(8)

21. Developing the promotion plan for thenew product (7)

22. Detailed design of the new product (11)23. Carrying out user tests (6)24. Finding and selecting suppliers (8)25. Building prototype (10)26. Testing prototype (10)

27. Building manufacturing facilities (8)28. Making technical drawings of the newproduct (11)

29. Developing tooling (10)30. Testing new manufacturing facilities (7)31. Debugging manufacturing process (5)32. Sales forecasting of the new product (6)33. Debugging the new product (10)34. Coaching the first production runs (8)35. Developing maintenance plans for the

new product (0)36. Training staff in assembling the new

product (5)

Type B:

Simple product & New project

3 (case C, case D, case E)

Type D:

Complex product & New project

3 (case I, case J, case K)

The familiarity

of the

NPD project

Type A:

Simple product & Familiar

project2 (case A, case B)

Type C:

Complex product & Familiar

project3 (case F, case G, case H)

The complexity of the NPD product

Figure 4. The 11 Cases Divided over the Four Categories





11/15

37. Testing promotion plan for the newproduct (1)

38. Developing maintenance documents forthe new product (7)

39. Training staff in installing the product (1)40. Promoting the new product (1)41. Training staff in maintaining the new

product (0)42. Developing the distribution plan for the

new product (3)43. Training users (7)44. Evaluating the NPD process (10)45. Certifying the new product (7)46. Patenting the new technology (6)47. Manufacturing the new product (0)48. Assembling the new product (1)49. Distributing the new product (0)50. Selling the new product (0)51. Installing the new product (0)52. Maintenance and servicing the new

product (0)

53. Recycling the new product (0)54. Eliminating the new product (0)

We see that seven different NPD activities arecarried out in all 11 cases. NPD activities thatare essential for all cases are:

8. Development of preliminary programmeof requirements of the new product (11)

11. Forming the project team (11)15. Preliminary planning for the total NPD

process (11)17. Carrying out the technical feasibility of the

new product (11)

18. Developing and testing concepts for thenew product (11)22. Detailed design of the new product (11)28. Making technical drawings of the new

product (11)

This seems to be a real minimal NPD process.You start with an NPD team, make a prelimi-nary plan for the project, start thinking aboutthe product requirements, generate concepts,check the technical feasibility of these con-cepts, start detailing the product design andfinally record everything in the technicaldrawings. It is indeed simple, it has an engi-

neering bias, but it could work! Probablybecause of the other level of abstraction, thisminimal process has more steps than theminimal process Cooper reported (seven stepsnow compared with only two from Coopersstudy).

Eleven NPD activities have not beencarried out in any of the cases (nos 1, 4, 35,41, 47, 49, 50, 51, 52, 53 and 54). This involvestwo pre-NPD activities (nos 1 and 4) andseven post-NPD activities (nos 47, 49, 50, 51,52, 53 and 54). These post-NPD activitiesinclude activities for manufacturing, sales,

maintenance and recycling, which are opera-tional activities the NPD team has normallynothing to do with. These activities areincluded in the theoretical list of 54, becauseduring design you have to be fully aware ofthese downstream aspects. Because the prod-ucts of our second empirical study wereavailable on the market (a must for being aparticipant in the Dutch design competition)these post-NPD activities must have beencarried out by the company, but were appar-ently outside the scope of the interviewees(NPD project leaders).

The two pre-NPD activities that were notcarried out (nos 1 and 4) are related to corpo-rate strategy, HRM and basic technologicalresearch. These are activities at a corporatelevel, which are indeed very influential forproduct innovation, but are also outside thescope of an NPD project leader.

The two core NPD activities that have not

been carried out in our cases are no. 35 Thedevelopment of maintenance plans for the newproduct, and no. 41 Maintenance training.Apparently, as in our first exploratory study, inour 11 cases the subject of maintenance seemsnot so important or is ignored.

Besides finding a minimalist NPD process,we were also able to find a kind of regularNPD process. We define regular here as theNPD process with the maximum numberof activities which are executed in at leasttwo cases in all four type categories (so theminimum score per activity should be 8).

The regular NPD process looks like this:

Appointing a project leader (10) Preliminary description of the idea for a

new product or a new challenge in themarketplace (8)

Development of preliminary programme ofrequirements of the new product (11)

Investigating the commercial feasibility ofthe new product (9)

Forming the project team (11) Determining and analysis of the target

group (9) Building and testing of experimental proto-

types (9) Generating and evaluating ideas for the newproduct (9)

Preliminary planning for the total NPDprocess (11)

Carrying out the technical feasibility of thenew product (11)

Developing and testing concepts for thenew product (11)

Analysing competitive products (9) Detailed design of the new product (11) Building prototype (10) Testing prototype (10)





12/15

Building manufacturing facilities (8) Making technical drawings of the new

product (11) Developing tooling (10) Debugging new product (10)

This regular NPD process has 19 activitieswhich represent a realistic product innovation

process. There is a project leader and a team, aproject plan, a product idea, requirements,generating of ideas and concepts, the targetgroup and the competition is investigated, both commercial and technical feasibilitystudies are carried out, the product design isdetailed, prototypes are built and tested, tech-nical drawings are made, tooling and othermanufacturing equipment is built and minorstart-up problems in the new product arecured. It looks like a balanced NPD process.

If we compare the results over the four cat-egories of NPD projects, we see some differ-ences (see Table 2). If we look at the number ofactivities that are carried out in all cases pertype category, we see that for the Type A cases,26 activities are executed in all cases, 10 activi-ties are carried out in only one of the cases and18 activities are not carried out. If we omit the10 NPD activities that have not been carriedout by any of the cases of all four types, youcould say that for the Type A cases 59 per centof all NPD activities is equal (26 of 44), and 41per cent is different. In the Type B cases, 15activities are carried out in all cases, five activi-ties are carried out in only one of the cases, soequality here is 34 per cent. For the Type C

cases, the number is 17 which give an equalityof 39 per cent. For the Type D cases the numberis 16, which is an equality of 36 per cent. Thenumber of activities not carried out decreasesfrom 18 for the Type As to 15 for the Type Csand Ds. This could suggest that Type A is lessdemanding than the other three. If we alsoinclude the equality over the types, then itwould suggest that Type A (Simple & familiar)is different compared to the other three typesof NPD projects. Is this marginal evidence

for the original suggestion of the experiencedproject leaders of the design agencies thatsimple products require a simple NPD project?

Results: Patterns in NPD Projects

We have now only analysed the existence ofcertain NPD activities (whether this activitywas executed in one of the cases). But it is alsointeresting to see whether we can identify dif-ferent patterns of activities. To do so, we havedivided the NPD process into sub-stages of20 per cent duration time each, and countedthe activities which were carried out in eachsub-stage. Activities which were present inmore than one sub-stage are counted as sepa-rate activities in all the relevant stages. Com-pared to the analysis in the previous section,we have many more NPD activities now, because if an activity is executed in threesequential sub-stages, it is counted three times.An overview of all the sub-stages over the fourtypes of NPD projects is shown in Table 3.

The Type As are doing the fewest NPDactivities and the Type Bs the most. Type Csare average and the Ds are doing a little more.It also shows that the As do little in the last twosub-stages and relatively more in the sub-stages 2 and 3. The Bs do a lot in the first andthe last stages. The Cs are above average insub-stage 1 and below average in sub-stages 2and 3. The Ds are above average in sub-stage 2,and below in sub-stage 5.

We also looked at the kind of sequences inwhich the NPD activities were carried out.Table 4 shows an overview. This table needssome extra explanation, because the score319, for instance for Type A in sub-stage 1,does not mean that all possible NPD activitiesbetween 3 and 19 were executed, but it givesan indication which kind of NPD activitieswere done in such a sub-stage. Thesesequences are derived from the theoretical listof 54 NPD activities (see Table 1). The empiri-

Table 2. Number of NPD Activities per Category

Type A Type B Type C Type D

In 3 cases Not relevant* 15 17 16In 2 cases 26 17 13 17In 1 case 10 5 9 6In 0 cases 18 17 15 15Total no. of NPD activities 54 54 54 54

* In the Type A category, only two cases are present.





13/15

cal data, however, show that inside thosesequences a lot of the activities are carried outin parallel. But no sub-patterns emerged. Itseems as though project leaders use a veryopportunistic attitude: they try to executeas many activities in parallel as possible. Theopportunities are neither type-based nor type-specific, but are mainly the result of the avail-ability of resources.

Based on the raw data, we see that the Type

As carried out 18 NPD activities in the firstsub-stage. Those 18 range from activity no. 3Exploratory market researchtono.19Analysingcompetitive products. Six activities wereexecuted in all Type A cases (= 12) and sixothers (makes the total 18) were carried outonly once, but not necessarily in one case.

The exceptional scores are labelled excep-tional because the activity was carried out at atime that was strange compared to its place onthe original overall list, which is followingmore or less the logical order of the NPDprocess. An interesting result in this respect is,

for instance, activity no. 32 Sales forecasting ofthe new product, which on the original list isplaced near the end of the NPD process, but intwo of the cases it was carried out during thefirst sub-stage. The same is true for activity no.42 Developing the distribution plan which isalso logical towards the end, but for one TypeD case (New & complex product) it was foundimportant to start thinking about the wayto distribute this new product quite early,

because that could influence the packagingdesign of the new product. This table is onlymeant to show different sequences of relatedNPD activities.

A tentative conclusion could be that the twotypes which score high on the newness axis(Types B and D) execute more NPD activitiesthan the other two. Looking to the NPDsequences shows that Types C and D (high oncomplexity) execute in the first sub-stage moreand different activities than the other twotypes. While the others start detailing thedesign (activity no. 22) in sub-stage 2, the two

Table 3. NPD Activities per Sub-stage for the Four Types of NPD Projects

Sub-stage in % Type A Type B Type C Type D Average

No. % No. % No. % No. % No. %

120 (sub-stage 1) 18 29 32 35 29 34 26 29 26.25 312040 (sub-stage 2) 17 25 12 13 13 16 20 23 15.50 194060 (sub-stage 3) 14 22 12 13 8 10 12 13 11.50 146080 (sub-stage 4) 7 11 14 15 14 17 16 18 12.75 1680100 (sub-stage 5) 8 13 22 24 19 23 15 17 16.00 20Total no. of NPD activities 64 100 92 100 83 100 89 100 82.00 100

Table 4. Sequences of NPD Activities

Sub-stage in % Type A Type B Type C Type D

120 319 31932*

22432*

32342*

2040 826 132837*, 46*

102845*, 46*

82432*, 37*

4060 203345*, 46*

932 1328 152945*

6080 2733 93240*

202943*

213845*, 46*

80100 3448 1646 2044 2145Total no. of NPD activities 64 92 83 89

* indicate exceptional NPD activities for that sub-stage.





14/15

complexity types start doing that already insub-stage 1. Once again, does this representsome small support for the original sugges-tion, that complex products need a complexNPD project?

Conclusion and DiscussionIn the preliminary theoretical study we distin-guished 54 different NPD activities thatdescribe the commonalities of 90 differenttheoretical models on NPD. According to thissummarized overview, the maximum NPDprocess can have 54 different activities.

In the first exploratory empirical study, weanalysed these activities. Most NPD projectsdo not use all theoretical activities to the full.Our four cases together covered nearly allactivities, but the two detailed cases describedused only 26 and 27 different activities. The

theoretical NPD models are much broaderthan NPD reality requires. Our minimal NPDprocess shows only seven different activities,whilst our regular process uses 19 differentNPD activities. These are rather smallernumbers than the theoretical 54!

We also discovered how experienced projectleaders from design firms categorize theirprojects in a very specific way before they planand organize them. They use the complexityof the product and the familiarity of theproject as main drivers to guide their plan-ning and timing of activities for future NPDprojects. Doing so, there are four categories ofNPD projects: Type A familiar + simple, TypeB unfamiliar (new) + simple, Type C familiar+ complex and Type D unfamiliar (new) +complex. Based on their experiences, the ideawas that simple and familiar cases need fewersteps and less duration time than the complexand new cases. They used this idea to plantheir future NPD projects.

In the main empirical study we checked thisplanning idea and also looked for more detailsbetween the four categories of NPD projects.Analysis of NPD activities and their durationtimes of 11 detailed cases of prize-winning

NPD projects executed within 11 differentcompanies shows interesting differences andcommonalities between the four categories.

We can conclude that a familiar NPD projecton a simple product (Type A) differs stronglyin the number of NPD activities that areexecuted in comparison to the other types ofNPD projects (Types B, C and D). In thisrespect it supports the original planning ideathat simple projects require a simple process, but the differences between the other threecategories are less clear and the results areinconclusive.

We have found a minimal NPD process,where you start with an NPD team, make apreliminary plan for the project, start thinkingabout the product requirements, generate con-cepts, check the technical feasibility of theseconcepts, start detailing the product designand finally record everything in the technicaldrawings.

We also found a process that we indicate asa regular NPD process, that has 19 activitiesand looks very convincing as a realisticproduct innovation process. There is a projectleader and an NPD team, a project plan, aproduct idea, requirements, generating ofideas and concepts, the target group and thecompetition is investigated, both commercialand technical feasibility studies are carriedout, the product design is detailed, prototypesare built and tested, technical drawings aremade, tooling and other manufacturing equip-ment is built and minor start-up problems in

the new product are debugged.On the patterns of NPD projects we can con-

clude that in all types of NPD projects mostactivities are carried out at the early stages ofthe project. NPD projects on new products(Types B and D) contain the most activities inthe total project, with a peak in the beginningand at the final stage of the project. ComplexNPD projects (Types C and D) execute evenmore activities in the first stage of the project,and also different activities from projects thatare not so complex. For instance, in complexprojects we found that detailing the designstarted already in the first stage of the project.We also found that NPD project leaders adoptan opportunistic attitude towards carrying outNPD activities in parallel in order to gain time.Their message is: if you can do activities inparallel, please do.

The results from these two empirical studiesprovide extra luggage for applying carefullythe well-known theoretical models of the NPDprocess. The experience provided by 11 suc-cessful project leaders may help novices tobuild their own expertise in a more profoundand reflective manner. Now the real actionplanning for NPD projects can begin! We hope

this exploratory study will give fuel to otherempirical studies on the NPD process in orderto balance the prescriptive NPD theorists,which will help NPD practitioners with theirdifficult work.

Acknowledgements

The author would like to thank Aernout vander Zee who executed the preliminary studyand did the original interviewing. RianneValkenburg, Marc Tassoul as well as the





15/15

anonymous reviewers of CIM made stimulat-ing suggestions for improving the text. Thisresearch was carried out in the Netherlandsduring 2003 and 2004. Reporting the resultswas done in 2006 and 2007.

References

Andreasen, M.M. and Hein, L. (1985) IntegratedProduct Development. IFS (Publications) Ltd/Springer Verlag, Bedford/Heidelberg.

Archer, B.L. (1971) Technological Innovation AMethodology. Inforlink, London.

Buijs, J.A. (1984) Innovatie en Interventie. Kluwer,Deventer.

Buijs, J.A. (1987) Innovatie en Interventie. 2ndupdated version. Kluwer, Deventer.

Buijs, J.A. (2003) Modelling Product InnovationProcesses, from Linear Logic to Circular Chaos.Creativity and Innovation Management, 12, 7693.

Buijs, J.A. and Valkenburg, R. (2005) Integrale Pro-ductontwikkeling. Lemma, Utrecht.Cooper, R.G. (1983) The New Product Process: An

Empirically Based Classification Scheme. R & DManagement, 13, 113.

Cooper, R.G. (1984) The New Product Process: ADecision Guide for Management. Journal of Mar-keting Management, 3(3), 23855.

Cross, N. (1994) Engineering Methods, Strategies forProduct Design, 2nd edn. John Wiley & Sons,Chichester.

Koen, P., Ajamian, G., Burkart, R., Clamen, A.,Davidson, J., DAmore, R., Elkins, C., Herald, K.,Incorvia, M., Johnson, A., Karol, R., Seibert, R.,Slavejkov, A. and Wagner, K. (2001) Providing

Clarity and a Common Language to the Fuzzy FrontEnd. Industrial Research Institute, Washington.

Pahl, G. and Beitz, W. (1984) Engineering Design: ASystematic Approach. Design Council, London.

Roozenburg, N.F.M. and Eekels, J. (1995) ProductDesign: Fundamentals and Methods. John Wiley &Sons, Chichester.

Saren, M.A. (1984) A Classification and Review ofModels of the Intra-Firm Innovation Process.R&D Management, 14, 1124.

Schrage, M. (2000) Serious Play. MIT Press,Cambridge.

Ulrich, K.T. and Eppinger, S.T. (1995) Product Designand Development. McGraw-Hill, New York.Van der Zee, A. (2003) Proceskeuze in de produc-

tontwikkeling. Unpublished masters thesis,Faculty of Industrial Design Engineering, Univer-sity of Delft.

VDI 2221 (1986) Methodik zur Entwicklung und Kon-struiren technischer Systeme und Produkte. VereinDeutscher Ingenieure, Dsseldorf.

Jan Buijs ([email protected]) is a professorin Product Innovation and Creativity at theFaculty of Industrial Design Engineering atthe Delft University of Technology. He has

been active for more than 30 years as ateacher, researcher and consultant. He hasworked with many different organizations:from low-tech SMEs to high-tech multina-tionals, from hospitals to insurance compa-nies, from consulting firms to universities.His work has ranged from helping start-upsto closing down mature industries. Hiscurrent research interests include multidis-ciplinary innovation teams and the relation-ship between branding and NPD. From2000 to 2007, he was chairman of the Euro-pean Association for Creativity and Innova-tion (EACI), a not-for-profit organizationwhich is bridging the gap between academ-ics and professionals in the field of innova-tion and creativity (http://www.EACI.net).

mailto:[email protected]://www.eaci.net/http://www.eaci.net/mailto:[email protected]

Documents

Action Planning for New Product Development Projects