winning businesses in product development

8/13/2019 Winning Businesses in Product Development

1/16

W I N N I N G B U S I N E S S E S I N P R O D U C TD E V H O P M B I I T T H E C R I T I C A LS U C C E S S F A C T O R SA formal new product process isn 'tenoughyouneed a high-quality process,

a clear a nd visible strategy,enoughpeople and m oney, and a respectableR&D budget. How does your program rate on these 10 metrics?

Robert G. Cooper and Elko J. KleinschmidtOVERVIEW: 2007 is Research-Technology Manage-ment s 5 0th year of publication. To mark the occasion,each issue reprints oneo/RTM s six most requently ref-erenced a rticles. The articles w ere identified by N.Thongpapanland Jonathan D . Linton in their 2004 studyof technology innovation management journals, a

Robert Cooper is professor of marketing at the M . G .DeGroote School of Business, McMaster University,Hamilton, Ontario, C anada, and president of theProduc t Developm ent Institute Inc. He is an activeresearcher, autho r and consultant in the field of productinnovation m anagement, and the developer of the Stage-Gateidea to launchprocess. He has won two MauriceHolland awards for the best paper published inResearch-Technology Management in 1990 ( NewProducts: What Distinguishes the Winners? ) and 1994( Debunking the Myths of New Product Development )[email protected]; www .stage gate.comElko Kleinschmidt is professor of marketing at the M . G .DeGroote School of Business, McM asterU niversity,an da prolific resea rcher and author in the field of manage-ment of new products. H e holds a mechanical engineer-ing degree from H anover (Germany) EngineeringCollege and an M.B.A. and Ph.D. from McG ill Univer-sity, Montreal.

citation-based study in whichRTM ranked third out o25 specialty journals in that field (seeRTM, May-June2004, pp. 5-6). The benchmarking study reprinted herewas originally published in 1996 and has been updatedwith its author's refiections. The ir study of 161 businessunits uncovered the key drivers of new product perfor-mance at the business unit level. Ten different perfor-mance measures were gauged including percentageof sales by new products, profitability and successrate. The ten gauges were reduced to two key perfor-mance dimensions profitability and impact whichdefined t h e performancem a p . Ninepossible driversincluding strategy, process, organizational design, andclimate for innovation wereinvestigated and four kedrivers of performance wereidentified:namely, a high-quality new product process, the new product strategyfor the business unit, resource availability, and R&Dspending levels. Merely having a formal new productprocess had no impact.KEY CONCEPTS:product development, new producperforma nce, critical success factors, strategy.What are the critical success factors that underlieexcellent new product performance? Our benchmarkingstudy was designed to uncover the drivers of perfor-mance by studying business units within corporations,and linking their new product performance to various

c ' c . ' h i i( > i o y v i y i ^ i i a g f e m ' e n t 0895-6308/07/$5.00 2007 Industrial Research Institute, Ine


2/16

practices. We measured performance in different ways,including profitability, success rates and percentagesales by new products. Som e ofthe business units studiedboasted superb performance; others had a more m ediocrenew product trackrecord By benchmarking against boththe good and the bad, we were able to obtain muchgreater insights into the factors and p ractices that reallydiscriminate between the top and poorer performers. Inshort, we were able to link specific practices to high per-formance.The results ofthisstudy of 6 businesses are provoca-tive. Here is a sample: The strongest driver of profitability is theexistence ofa high-quality, rigorousn wproduct processone thatemph asizes up-front homew ork, tough Go/Kil decisionpoints, sharp early produc t definition, and flexibility. Bycontrast, merely having a formal new product processhas no impact at all on performance The roleof new product strategyin the business unitwhether or not there is one, what it contains and whetherit is clearly communicatedhas a pronounced effect onperformance. Resourcesboth people and moneyare stronglytied to new product performance. R&D spending per sehad a strong but very focused im pact: It drove only som eperformance measures, butnot profitability. The use of cross-flinctional team s helps, but does nothave the dramatic impact on performance that we hadexpected. Rather, the quality ofthe teamsseems to makeall the difference to the business u nit's new product per-formance.Before discussing the critical success factors in detail,here is some background to the study: its rationale, itsframework, and what was measured and how.In Search of the Critical Success FactorsThe quest forthedrivers of new product success has beena popular research topic for the last decade or so.However, merely studying successful and unsuccessfulnew product projects, as we too have done in the past,misses many ofthe key factors in success. Consider this:The typical study, which focuses onspecific new productprojects, compares and contrasts the characteristics ofwinning versus losing projects, and in so doing, uncov ersthose factors that discriminate between the two (see, forexample,1,2 .This success/failure research has uncovered manyfactors that we now take for granted as driving success(see Table 1). But this research does have its limitations.Perhaps the most fundamental flaw is the fact thatresearch done at the project level often fails to identify

May^Junc 2 7

those company practices that decide success. There arethree reasons:1. First, success at the company or business-unit levelmay differ from success at the project level. For example,a business may have a string of "successful" projects, asmeasured by retum-on-investment of each project; butoverall, these projects have relatively little impact on thefirm, and so the business's total new product effort isjudged to be mediocre.2. Next , of ten importan t pract ices for exam ple,creating an innovative climate and cultureare notreadily apparent or measurable at the product level. Con-sequently, they are missed in such success/failurestudies.3 .Amore subtle and final reason for omissions is the waythe research is designed. When pairs of successes andfailures are selected from each firm, com pany character-istics that may have a strong impact on success will becommon to both projects. And so these company charac-teristics do not em erge in an analysis of factors that dis-tinguish the successes.In recent years, benchmarking against other companieshas become popular, and has in part overcome some ofthe deficiencies listed above. But the benchmarkingpracticed by most firms also has its problem. First, it isvery time-consuming; second, there is the problem ofgetting the cooperation of other firms; third, there aremethodological problems: the fact that those in thecompany doing the benchmarking may not be experi-

Table 1 Factors hatDrive New Product Success at theProject LevelStrategic Factors:Product advantageMarketing synergyTechnological/Manufacturing synergyAvailability of resourcesStrategy of the new productDeveiopment Process Factors:Proficiency of technological activitiesProficiency of marketing activitiesProficiency of up-front (homework) activitiesTop management supportSpeed to marketProficiency of financial/business analysisMarket Environment Factors:Market potential/sizeMarket competitivenessExtemal environmentOrganizational Factors:Internal/external relations (of team )How team was organizedSource: Montoya-Weiss and Calantone (2).


3/16

enced at such research , are not sure what to me asure, talkto the wrong people, interpret the data incorrectly, andso onin short, all the classic errors made by an inexpe-rienced investigator.

Finally, relationships typically are not studied. Bestpractices are noted, but the link between practices andimproved performance remains largely speculative.Thus,there always remains the concem that an observed

How the Research Was D oneNumerous studies into new product success and failure,together with case studies and popular books, havesuggested myriad factors that ought to drive new productperformance. From these, we developed a framework,which identified five major blocks of characteristics formeasurement:1. Process: the firm s new product development processand the specific activities within this p rocess.Processesthat promote a strong market orientation, undertaking themarketing tasks in a quality fashion, doing the pre-development activities well, and having sharp, earlyproduct definition, are purported to yield positiveoutcomes.2. Organization: the way projects are organizedInvestigations conclude that the use of cross-functionalteam, interfaces between departments, and an empoweredleader yield much better results than a functionally basednew product effort.3 . Strategy: the firm s total new product strategy.Having an explicit new product strategywhich definesthe role and goals of new product development in thecompany's overall strategy, specifies product/marketarenas as areas to focus on, and formalizes the necessaryorganizational structures for implementationresults inmore positive performance.4. Culture: the firm s internal culture and climate forinnovation.Facets of positive climate include: encour-aging intrapreneurship; providing support (rewards, risk,tolerance, autonomy, and acceptance of failures withoutpunishment); fostering the submission of new productideas;and providing free time and resources to undertakecreative activities.5. Com mitment: senior management s involvement withand commitment to new product development.Successfactors here include: senior management commitment torisk-taking; clear messages from senior managementabout the importance of new product development; andavailability of funds and resources for product develop-ment.Data were collected via a detailed, four-page question-naire. The questionnaire contained 48 measures thatcaptured new product practices (based on the five blocks,outlined above). These 48 measures were gauged via 1-5Likert-type scales with anchor phrases, and were used todevelop eight main themes or constructs (via addingsubsets ofthe 48 variables together). Cronbach alphas anditem-total correlations were used to check for internal

consistency of these constructs. R&D spending (% ofsales) and proportion going to new products were alsogauged.Ten performance merits were measured; eight of thesewere captured on 1-5 Likert-type scales, again withanchor phrases; two were directly measured as percent-ages (success rates and percentage sales by new products).These ten performance metrics were reduced to twounderlying performance dimensions via factor analysis(SPSSX routine; principal component analysis; varimaxrotation). The drivers of performance were identified bycorrelating the eight constructs (which captured newproduct practices), R&D spending, and the 48 character-istics (mentioned above) versus the two performancedimensions. Additional characteristics ofthe company ordivisionsize, R&D spending, location and industrywere also measured. The questionnaire itself was exten-sively pre-tested via personal interviews.Firms were selected from private lists of companies,compiled from databases and directories of companiesknown to be active in new product developm ent. A total of161 business units from Europe (primarily Germany andDenmark) and North Am erica (U.S. and Canada) partici-pated in this study. Questionnaires were directed tocorporate executives responsible for their business un it'snew product development effort. All respondents hadmore than three years involvement here; many had con-siderably more experience. Follow-up personal inter-views were conducted with particularly proficient firms toidentify best practice s and provide d the anec dotalevidence and examples cited. For more detail, see {3,4).The average annual sales revenue for the business units(or divisions) was US$562.8 million, with no significantdifferences between North Am erican and European firmsin the sample. R&D spending was 7.5 percent of sales onaverage, with 71.8 percent of this going to product devel-opment. No significant differences emerged betweenNorth American and European firms in terms of theseR&D spending pattems. The industry breakdown was:

PercentChemicals/materialsCommunication productsMachinery/equipmentElectronics/electrical productsFoodAutomotive parts/componentsMiscellaneousTotal

Research R G C

Technology

12.611.933.31.59.63.08.1100andE J KManagement


4/16

practice does not really have much impact on perfor-mance.The current study sought to overcome both sets ofproblems. First, by focusing on the business unitas theunit of analysis, rather than individual new productprojects, we identified a number of company character-istics or practices often missed in traditional research.Second, the research is rigorously designed, employs alarge sample of firms, and utilizes sophisticated dataanalysis methods. Finally, thelinkbetween practice andperformance is investigated, so that one can concludethat certain practices really do discriminate top perform-ers from the rest.A total of 6 business un its in a variety of industries inthe United States, Germany, Denmark, and Canadaagreed to participate in the benchmarking study. A diag-nostic questionnaire wa s directed to the senior executivein charge of product development in the business. Thequestionnaire was designed to characterize the busi-nes s's new produ ct practices in each of five major areas:process, organization, strategy, culture, and com mitment(see How the Research Was Don e, previous page).These five areas were measured via 48 metrics, fromwhich nine themes or dimensions of firms' practices andcharacteristics were identified (listed in Table 2). Inaddition, ten performance metrics captured how w ell thebusiness uni t's total new product effort performed.The Performance MapTen key performance metrics were gauged:o Success rate The proportion of development projectsthat became commercial successes.o Percentage of sales by new products The percentageof the business unit 's sales accounted for by newproducts introduced within the last three years.

o Profitability relative to spending How profitable thebusiness uni t's total new product efforts we re, relative tothe amount spent on them.o Technical success rating How successful the totaleffort was from a technical/technological perspective.o Sales impact How strong an impact the total newproduct effort had on the business uni t's top line or salesrevenues.o Profit impact How strong an impact the effort had onthe business unit's bottom line or annual profits.o Meeting sales objectives The extent to wh ich the totalnew product effort met the business unit's sales objec-tives for new products.o M eeting profit o bjectives The extent to which it metthe business unit's profit objectives.o Profitability versus competitors How profitable thetotal new product effort was relative to competitors.o Overall success All things considered, how success-ful the business unit's total new product efforts werewhen compared to competitors.Overall, our sample of companies fared quite well interms of the performance of their total new product

Table 2. The Impact of Nine Drivers of New Product Performance at the Business Unit Level

Performance DriverA high-quality new product processA defmed new product strategy for the business unitAdequate resourcespeople and moneyfor new productsR&D spending on new products (as % ofthe business's sales)High-quality new product development teamsSenior management commitment to new productsAn innovative climate and culture in the business unitThe use of cross-functional team s for product developm entSenior management accountability for new product results

Effect onProfitability^ .416.228.244ns .196^ .268.243.230.228

Num bers are correlation coefficients between specific drivers and two performance dimensions. All significant at the 0.01 level. Boldarrows denote strong co rrelations, significant at the 0.001 level.

May-June 2007 | ^ |

Effect onImpact.226.211.197-> .395.208nsnsnsns

numbers and


5/16

efforts when gaug ed on these ten metrics. New productsintroduced in the previous three years represented 28.4percent of annual sales, on average; 56.9 percent ofprojects that entered development were eventuallylaunched as commercial successes; and on the eightscaled performance metrics, companies scored aboutmid-scale, on average.A Performance Map was constructed based on these tenperformance metrics. A review of performance madeclear that the ten m etrics were strongly connected to oneanother. Therefore, factor analysis was used to identifythe underlying dimensions of performance and todevelop a Performance Map (SPSSX routine, principalcomponent analysis, varimax rotation; two factors wereselected based on both the Scree test and eigenvalues >1.0). Two performance dimensions emerged:1.Profitability This was the strongest dimension, andcaptured how profitable the businesse s s total newproduct efforts were. It comprised: the profitability

versus com petitors and overall success rating of the busi-ne ss s total new product effort; whether the total initia-tive met product objectives; its profitability relative tospending; and the impact of the total effort on thebusiness unit s profits.2.Impact This performance dimension, also a strongone, had less to do with profitability and more with theimpact that the total new product efforts had on thebusiness. It comp rised: percentage sales by new productsachieved by the business unit; the impact of newproducts on both sales and profits of the unit; the su ccessrate achieved; and the technical success rating.The two dimensions can be used to capture the newproduct performance of business units in a macro orbroad way; thatis , new product performance measured invarious ways boils down to two fundamental or underly-ing dimensions, namely profitability and impact on thebusiness These two dimensions thus becom e the verticaland horizontal axes of the Performance Map in Figure 1.

1.0

0.5ooI O

in

I

7.0

ay Factor loadings - upper right quadrant4=Tochnlcal success rating7=Meetlng Business s Sales Objectives

. 8=Meetlng Business s Profit Objectives9=PromabllltyVs. Competitors10=Overall Success Vs. Competitors

1

4 J j 3=PromrelativeJ ff J to pendingI / / / ,7 y6=ProntImpact/ / / / / 5^,e,,n,pact

/ / /A//\^^^^Sjr^=S cessnato

2=Percent Sales byNew Products

11.0 0.5.5 0.0

Impact on the Business UnitFigure 1 ThePerformance Map is based on the 10 performance metrics The arrowsshow ho w each metric loads on the two map dimensions Impact and Profitability

1.0

Research Technology anagement


6/16

Here we also show the ten performance metrics,revealing visually how each m etric loads on each perfor-mance dimension (the correlations between each perfor-mance metric and the two performance dimensions).Four Key Su ccess FactorsWhat then are the critical success factors in productdevelopmentthe factors that drive performance at thebusiness unit level? The study uncovered nine factorsthat distinguished the better performing businesses, fourfactors in a very strong w ay. The top four are:1. A high-quality new product process.One thatdemanded up-front homework, sharp and early productdefinition, tough Go/Kill decision poin ts, and quality ofexecution and thoroughness, yet provided fiexibility.2. defined new product strategy or the business unit.One in which: There were new product goals for thebusiness unit; areas of focus w ere delineated, the role ofnew products w as clearly com municated, and there was alonger-term thrust.3 . Adequate resources of people and m oney.Wheresenior managem ent had provided the needed people (andfreed up their time for projects), and resourced the effortwith adequate R&D funding.4.R D spendingfor new product developm ent (as a per-centage of sales).The other success factors, with a more modest effect onperformance, included:5.High-quality new product project teams.6. Senior management committed to, and involved in,new products.7.An innovative climate and culture.8.The use of cross-functional project teams.9. Senior management accountability for new productresults.Table 2 summ arizes the effects of these nine drivers (thecorrelations between the success factors and the perfor-mance dimensions), while Figure 2 shows the Perfor-mance Map, but this time indicating the direct linksbetween performance and business practices. Considernow the top four success factors in more detail:1. A high-quality new product process.This wasthe s t ronges t common denominator among high-performance businesses (see Table 2 and Figure2).Herethe term "new product process" means those steps,activities and decision-points that new product projectsfollow from idea to launch and beyond.

A word of caution here: The mere existence oi a . formalproduct development process had absolutelyno effectonperfonnance; there was no correlation at all betweenmerely having a process and performance results. Sothose companies that mistakenly believe they can "gothrough the motions" and re-engineer their new productprocesses (usually amounting to documenting what theyare already doing ) are in for a big disap pointm ent.Having a process did not seem to matter; rather it wasthequality and nature of that processbuilding in bestpracticesthat really drove performance.Here are the six ingredients we measured to gauge thequality of the process, and when co nsidered together, addup to a quality process and positive performance. Wealso use the anecdotal evidence from the study to showhow some firms were particularly proficient in terms ofthese six ingredients: First, there was an emphasis on up-front homeworkboth ma rket and technical assessments before projectsmoved into the development phase.Too many projects move from the idea stage right intodevelopment with little or no assessment. The results ofthis "ready, fire, aim" approach are usually disastrous.Inadequate up-front homework has been found to be amajor cause of failure in product development. In thebest new product processes we saw, management haddeliberately built in one or two homework stages beforethe "go-to-development" decision point, comprisingvital "must do" actions. For example, in the five-stageprocess employed by the Professional Division of SCJohnsons Wax, the two homew ork stages entail:

May June 2007


7/16

0.5

0.4SsUsaofCrost Funetlonalreams8sSn/orManagmmnt ccounbblUty

HlghQuality Naw Product Procaaa

Sanlor Managamant CommHmantBAdaquataResources

aDaflnadNPStratagyQualityTeams

(Upper right quadrant ofPerformance Map)

(0.1)

2sR &D Spefiding on New Producti( of Sales)I

0.1 0.2 0.3 0.4 0.5Correlations with Impact

Figure 2. How the critical success factors drive new product performance as gaugedby correlations of nine drivers w ith thetwo performance dimensions.

Preliminary market assessment: A preliminary andquick assessment ofth e market potential, need level, andcustomer requirements, done very early in the project. Preliminary technical assessment: A parallel activityto identify the technical possibilities and risks. Detai led market s tudies: User needs-and-wantsstudies (to pin down exact customer need s, requirementsand benefits sought), competitive analysis, and concepttesting (to ascertain probable purchase intent). Detailed technical assessment: Determination of theprobable technical rou te, risks, patent position, m anufac-turability, costs and capital requirements, timing andresources required.

Financial and business analysis: Based on the aboveactions, a profitability analysis (a discounted cash flowwith sensitivity analysis) and the business rationale formoving ahead with the project.The result of these homework actions at SCJ is abusiness case a mandatory element prior to movingthe project into the development stage. heprocess included sharp early product definitionbefore development w ork began.Failure to define the productits target market; theconcept, benefits and positioning; and its requirements,features and specsbefore development begins is amajor cause of both new product failure and serious

RL search Technology Management


8/16

delays in the development cycle. Some companies, suchas Hewlett-Packard, hav e placed major emph asis in theirPhase-Review Process on getting the product defmitionpinned down before a formal development project isapproved. This defmition, of course, is based on facts,rather than hearsay and specu lation; hence the need for asolid up-front homework phase.o There were tough Go/Kill decision points in theprocess, where projects really did getkilledProjects tend to take on a life of their own In too man ycompanies we investigated, projects moved too far intodevelopm ent w ithout serious scrutiny. And it was only asthe project approached commercialization that the hardtruths were recognized: The marketw snot quite as largeas expected, or manufacturing costs were higher thananticipated, and so on. The lack of tough Go/Killdecision points meant too many product fai lures,resources wasted on the wrong projects, and a lack offocus.The result was too many marginal projects in thepipeline, while the truly meritorious projects werestarved.In the better processes we observed, companies haddesigned funnelling or culling process in the form oftough review points orgates.For example, in Rohm &Haas's WIN process, there are five gates, where seniormanagement reviews the deiiverables, evaluates theproject against pre-set criteria, and approves the actionplan and resources for the next stage. The decisionprocess is timely and efficient, and ensures that poorprojects are weeded out before excessive spendingoccurs.o There was focus on quality of execution, in whichproject activities were carried out in a quality fashion.An emphasis on quality-of-execution in many firmscame abou t after internal studies revealed that too manyprojects suffered from weak, inconsistent work, some ofthe most deficient areas being the market-related ones.Top-performing firms work at improving quality ofexecution of key tasks and activities throughout theprocess, from idea generation right through to launch.Some companies, such as ECC in Atlanta (the world'smajor produce r of clay) have re-engineered their productdevelopment processes in order to achieve quality ofexecution improvements. By specifying the key deiiver-ables at each gate or decision point, and by conducting athorough review at the gate, (where the quality of workdone in the previous stage or phase is rigorously scruti-nized), the quality of work has significantly improved.Project teams and leaders know what rigor is expected ofthem, and hen ce set their own quality or action standardshigher.

o The new process was complete or thorough; everyneeded activity was carried out, with no hasty corner-cutting.Many companies had discovered that not only was thequality of work lacking, but in some ca ses, thework waslacking altogether. That is, key tasks, such as marketanalysis, business assessment, and customer research,were simply not done (or left until far too late in theprocess). This deficiency caused some companies toredesign their processes, building in these tasks at theappropriate point in the process.In ECC's process, deiiverables have been defined apriori for each decision-point or gate review; these deiiv-erables specify what is required at a given point in theproject, and hence determine what work or tasks must beundertaken within a given stage.o The new product process was flexible; stages anddecision points could be skipped or combined, asdictated by the nature and risk oftheproject.One pitfal l some firms encounter when they dore-engineer their product development process is thefailure to build in flexibility. Instead of being a temp lateor roadmap, the "formal process" becomes a straight-jacket beset with bureaucracy. By contrast, companiessuch as Procter & Gamble specify in their new productprocess that flexibility is key, that certain steps andactivities can be omitted, and that the beginnings andends of succeeding phases overlapped "provided that weunderstand the risks involved and have agreed at theprevious decision checkpoint" (5).

May -June 2007


9/16

A high-quality new product process clearly pays off.Companies tha t boas ted p rof ic ien t new produc tprocessesones that incorporated the six ingredientslisted abovewere rewarded with superior perfor-mance: higher profitability and higher impact efforts.Once again, note that merely ha ving a formal proc ess had

no impact; it wasthe qu ality of that processthat made thedifference2. defined new product strategy or the business unit.Having a clear and visible new product strategy was theNo. 2 driver of businesses's new product perfonnance.

"W inning Businesses in P rodu ctDevelopment: The Critical SuccessFacto rs" RevisitedWhy do some businesses make product innovation seemso easya steady stream of new product winners? Our1996 article reported the results of our first major bench-marking study that investigated new product development(NPD) performance results and how top-performing firmsachieved positive results.A number of best-practice themes emerged from that 1996study, and they are as relevant today as then (since then,we have undertaken other benchm arking studies in orderto refine our insights into what leads to success in productinnovation, also reported in RTM, 1). The outcome ofthese studies is theperformance diamond, next page,which summarizes the four major performance drivers ofNPD results (2):1.StrategicTop performers possess aproduct innova-tion strategy, driven by the leadership team and itsstrategic vision for the business. Notably, even today,about half of businesses lack key facets of this strategyThis innovation strategy consists of a number of elem ents,including the business's goals for product innovation (forexample, 3M's goal of "percentage of sales from newproducts" has been adopted by m any firms) and how thebusiness's new product effort ties into its overall businessgoals.Arenas of strategic focuswhere the business willfocus its R&D effortsare also a part of the innovationstrategy, along with how the business plans to win in eacharea.Attack plans include strategic stance, entry strategy andalliance s trategy (for example, P&G's s trategy of"connect and develop" or working with partners todevelop new products outside the corporation). And theinnovation strategy includes the product and technologyroadmaps which spell out the major development initia-tives (for exam ple, HP maps out its major developmentsover a five-to-seven-year time horizon).2 . Portfolio manag ement and resource allocation Asecond common denominator of top-performing busi-nesses is making sure that the business has the necessaryresources available for NPD, both funds and peoplefrom all functional areas. But deep pock ets is not the onlydriver of high performance; rather, astute investment ofthese resources is keytoo:top performers have ^portfolio

management system that helps the leadership teamallocate these resources to the right areas and rightprojectsthe right mix and balance of NPD investments,and a strategically aligned portfolio.Strategic buckets is a method designed to allocateresources to the right strategic arenas and to achieve theright balance of projects 3) .Note, for exam ple, that best-performing businesses have a different breakdown in thetypes of NPD projects, with a much higher proportionof bolder, larger and riskier ventures than do poor per-formers 4) .Top-performing firms also maximize the productivity oftheirR Dspending: they ensure that funds and people a refocused on high-value projects. The best firms rely onseveral powerful methodsscorecards, the productivityindex, and real optionsto select and prioritize their newproduct projects (5).3. The NPD ProcessMost firms have implemented agating or Stage-Gate idea-to-launch new productp roces s 6) . But there is great variability amongcompanies in how well the process works. The top per-formers have a well-crafted, robust new product processin place , one that drives NPD projects from idea to launchand beyond ; the i r p roces s em phas iz es up - f ron thomework, voice-of-customer input, quality of execution,and performance results m etrics.Top performers embrace their NPD process and practicediscipline in its implementation; they have also stream-lined the process in the last ten years, making it flexible,adaptive and scalable (for example, there are Lite andXPress versions of Stage-Gate for lower-risk projects).Finally, they have built lean principles into their NPDprocess (7).4. People: culture, climate, teams, and the role of seniormanagementA fourth dimension was later added to the1996 study's results to yield the performance diamonddepicted on the following page. This fourth dimensionis about people. In top-performing businesses, there isa positive climate and culture for innovation; the leader-ship team of the business actively supports innova-tion with words, actions and resource commitments,and senior management is engaged in the decision-making process in the right way. Additionally top-performing businesses embrace a true team approachto NPD : they f ield ef fect ive , prope r ly- resou rcedcross-functional teams that are accountable for the endresult.


10/16

Strategy was linkedtobo th performance dimensions (seeFigure 2 and Table 2).Here are the four main ingredients of a positive newproduct strategy, which when taken together, add up topositive performance:

There are goals or objectives for the business s totalnew product effort e.g., whatsales,profits, e tc. the newproducts will contribute to the business).What surprised us was how many businesses lackeddefined, .clear and written go als for their to tal new -

The diamond had its seeds in our 1996 RTM article,reprinted here, and captures the key success drivers inproduct innovation. Use it as a compass or guide tomanaging your innovation effort; for example, Procter &Gamble is one leading firm that has adopted thisdiamo ndthey call it the Initiatives Diamond and ithas proved to be a powerful guide to product innovationthere {8).R.G.C. and E.J.K., May 2007.

Notes and References1. Subsequent benchmarking studies are reported in the three-part RT M series: R. G.Coo per. Edgett, S. J. , and Kleinschmidt E. J. Benchm arking Best NP DPracticesParts I, II and III. Research-Technotogy Management, 47, 1, Jan .-Fe b. 2004,pp. 31-43; also :47. 3. May-June 2004, pp. 50-60; and 47. 6.Jan . -Fe b , 2005 , pp . 43-55 ,

2. More on the perfo rmance d iam ond in R, G, Cooper , Product Leadership: Pathwaysto Profitable Innovation, 2nd edition. New York, NY: Perseus Publishing, 2005.www.stage-gate.cont3. For mo re on strategic buckets see the two-part RT M series: R. G. Cooper, Edgett,S. J. and Kleinschmidt E, J, Portfolio ma nagem ent in new product develo pment:lessons from the leadersParts I & 11. Research-Techno logy Managem ent 40, 5,Sept.-Oct. 1997, pp. 16-28; and: 40, 6, Nov.-Dec, 1997, pp, 43-52.4. Coo per , R. G, Your NP D por t fo l io may be harmful to your business 's heal th . PDMAVisions. XXIX, 2, April 2005, pp. 22-265. For project selection methods, see: R, G. Cooper and S, J, Edgett. 10 ways to makebetter portfolio and project selection decisions. Visions Magazine (PDMA), XXX, 3, .June 2006, pp, 11-15.6. Stage-Gate is a registered tradem ark of the Product Deve lopme nt Institute Inc.7. R, G, Coo per and S. J. Edgett. Lean. Rapid and Profitable New ProductDevelopment. Product Development Inst i tu te, www.stage.gate.com 2005.8. The P&G Initiatives Diamond and how it is used to achieve superior NPD results isreporte d in R. G. Coo per and M, Mills, Sueceeding at new pro ducts the P& G wa y: Akey element is using the Innovat ion D iamond, PDMA Vision.t, XXIX, 4, October2005, pp . 9 -13,

Business s newprodu tperform n e

Idea-to-:tau.n ^h.jSysti^m: -iT-ilS t a g e - ^a t e

The Performance Diam ond dep icts the four main drivers of positive NPD performance.Use it to guide your NPD efforts Slage-Gale/.va trademarko fProduct Development Institute Inc.)

May-June 2007


11/16

product effort. And thelacicof ttiis rather basic ingredientof strategy had negative consequences for the business.By contrast, leading firms, such as 3M, make newproduct goals such as "30 percent of our division 's saleswill come from new products introduced over the nextthree years" an explicit part of every division's businessgoals. Other commonly cited goals, besides percentageof sales, are "dollar sales to be generated from newproducts"; "percentage of profits"; and "numbers ofmajor and minor launches per year." herole of new productsinachievingbusinessgoals isclearly communicated to all.The reason for having goals is that everyone involved innew products can have a common purposesomethingto work toward. Far too often, the people w ho worked onnew product projects w ere not aware oftheir business'snew product objectives, or the role that new productsplayed in the total business objectives. The message isthis: Do what 3M does. Set goals for the new producteffort (e.g., percentage of sales, profit or growth over thenext years), and mak e them cleartoeveryone involved. There are clearly defined areas of strategic focusstrategicarenastogive direction to the business's totalnew product effort.The new product strategy specifies "the arenas where wewill play the game," or perhaps more important, wherewe won't play . . . what isin boundsandout of bounds.These arenas w ere often defined in terms of the types ofproducts, markets or technologies the business unitwould focus on. With arenas undefined, the search forspecific new prod uct ideas or opportunities is unfocused;over time, the portfolio of new p roduct projects is likelyto contain a lot of unrelated projects, in many differentmarkets, technologies or product-typesa shotguneffort. The business's new product effort has a long-termthrust and focus, including some lon ger-term projects(as opposed to short-term, incrementalprojects).The short time horizon of firms has been a criticismwidely voiced. Importantly, this one ingredient was themost important ofthe four strategy ingredients, and sig-nificantly linked to a number of specific performancemetrics.A sound new product strategy lies at the heart ofabusi-ness's new product effort. Those businesses that lackedgoals for their total new product effortswhere arenasor areas of strategic thrust had not been defined, wherethe strategy and projects were short-term in nature, andwhere the strategy was not well-comm unicated were ata decided disadvantage on both performance dimen-sions.Corning Incorporated, for example, has heeded themessag e. The com pany is installing a first-rate portfolio

management approach to help allocate resources to theright projects within business u nits. The approach does asuperb job of linking the project portfolio to strategy, andbegins first with a clear delineation of the business u nit'sstrategy: its goals for new products, and the arenas ofstrategic focus.3. Adequate resources of people and money.Top-performing firms had in place the needed resources toundertake new products; that is, senior management hadmade the necessary resource commitment, and kept iThis was the No. 3 driver of superior performance.Resource com mitment drove both the profitability of thebus iness 's total new product effort, as well as the impactof this effort on the business (see Figure 2 and T able 2).Three main ingredients lead to resource adequ acy and, inturn, positive performance for the business: The necessary resources have been devoted by seniomanagement to achieve the firm's new product objec-tives.In an effort to overcome weak new prod uct performance,some managements had undertaken a re-engineering orstrategic planning exercise. The problem was that theresulting strategy, goals or processes w ere not backed upwith the needed resources. And so the initiative floun-dered. As one manager eloquently stated, "Even the bestgame plan in the world comes to nothing if there aren'tplayers on the field "A lack of resources continues to plague new productprojects, and was often the culprit underlying poorquality of execution; there simply weren't the necessarypeople in place, nor the time available to do a qualityjob.The result was that comers w ere cut, activities were donein haste, tasks were left outand the results were pre-dictable. The point is: For positive results,the resourccommitment must be aligned with the business's newproduct objectives and processes. R D budgets are adequate to achieve the stateobjectives.This is just another facet of the resource question, butwith a specific focus on technical bud gets. Later, we lookspecifically at R&D spending as a percentage of sales,and its effect on new product performance. The necessary people areinplace, and release time given for specific new product projects.Projects are approved, and people are assigned to them.Too often, the assigned people are expected to work onanother six projects, or in the case of marketing andmanufacturing p eople, to do "their realjob"in addition tothe new product project. Some enlightened firms,however, are taking steps to overcome this deficiency. AtAlcoa's Knoxville packaging division, for example,when a project's action plan is approved at each gate

Ui'Sftirch I't'chnolog^' M Jan agc me nt


12/16

review, the plan spells out resource commitments,people and their t ime al location. Assignments ofpersonnel to specific projects are m ade realistically, andin full awareness oftheir other duties and obligations.Having a solid new product process (critical successfactorN o. 1) is only part of the answer. So too is havingan enunciated new product strategy (factor No. 2).Unless the process and the strategy are properlyresourced with people, time and moneyand the com-mitments keptdon't expect stellar performance.4.R D spending.R &D spending for product develop-ment, measured as a percentage of sales,was by far thestrongest determinant of theimpactof the product d evel-opment effort (see Table 2 and the horizontal dim ensionin Figure2).But R&D spending hadno significant effectwhatsoever on the other performanc e dimen sion;namely, profitability of the total new product effort.Thus, metrics that capture performance magnitude orimpact, such as the popular, "percentag e ofsalesby newproducts" measure are driven by the magnitude ofspending.The me ssage is clear: If your performance goal is to havea high-impact new product effortfor exam ple, toachieve a high percentage of your business unit's salesfrom new productsthen increased R&D spending isthe most obvious lever to pull. The relationship is notone-to-one , but it is strongFiveOther Su ccess FactorsAnother five drivers of new product performance werealso identified (see Figure 2), but their impacts weremore modest than the four critical success factors high-lighted above.5.High-quality new product teams.T he various orga-nizational and team measures contained two distinctfactors. The first captures the quality of the team, w hichwas a fairly strong driver of performance; the second(No.8, next page), whether it was a true cross-functionalteam or not, played a lesser role. The first of these, ahigh-qualityteam,meant:o The team leaderwas dedicated to thison project (asopposed to trying to lead many projects, or having manyother assignme nts). Sadly, in too many business units westudied, team leaders were spread too thinly across toomany projects or had too many other duties to runprojects effectively.o The team interacted and comm unicated well and ofiten,W\t\i frequent project update meetings,progress reviews,and problem resolution sessions. The best teams wewitnessed had short but weekly meetings to ensure thatthe entire team was up to speed.o Decisions made by outsider groups or people (outsidethe team) were handled quickly and efficiently. This

was usually the result of proficient team actions.For example, the team was able to do whatever inter-nal marketing, communication and persuasion wasnecessary to get outsiders on-board and to deliver quick,efficient decisions.Firms that boasted good-quality projectteams,as definedabove, had positive results on both performance dimen-sions (see Table 2 and Figure 2).6. Senior management commitment.Businesses withgreater senior management commitmentto,and involve-ment in, new products boasted more profitable total newproduct efforts. Firms where senior managements werevery much involved in new products were ones in which:o Senior management was strongly committed to newproducts and product development.o Management had committed the necessary resourcesto achieve the firm's new product goals.o Senior management was closely involved in theproject Go/Kill and new product spending decisionsthey had a central role in the new product project reviewprocess.7. Innovative climate an d culture.The climate forinnov ation w ithin the firm was a success factor, .butperhaps not as strong as one might have exp ected. S adly,the average scores on individual ingredients here wereamong the lowest of any success factors. Here, positiveclimates were ones in which:o There was a new product idea scheme within thebusiness unit, which solicited ideas from all employees.o Technical people w ere given rge time, scouting timeor time off to work on projects of their own choice.Typically this was between1 and 20 percent of the workweek. Here we saw some evidence that, while som e firmsmade this option available to their technical employees,relatively few employees took advantage of the offer.o Resources were made available to employees so thatthey could informally advance their own projects orundertake creative work of their own choice. Such

May June 2007


13/16

resources often included seed money for technicalresearch, bootstrapping accounts to fund unapprovedprojects, etc. Skunk works or teams working on unofficial projectswere encouraged.8. Cross-functional team.This second facet of theproject team captured whether or not it was a true cross-funct ional team. Here, posi t ive performance wasachieved when: All projects had a defined and accountable teamleader a person who w as responsible for advancing theproject. Project leaders were responsible for the project frombeginning to end (as opposed to being responsible for

The Cornerstones of New Product Performance1. A high-qualitynewproduct process meaning: An emphasis on up-front homew orkboth m arketand technical assessmentsbefore projects move intothe development phase. Sharp, early product defmition, before developm entwork begins. Tough Go/Kill decision points in the process , whereprojects really do get killed. A focus on quality of execu tion, where activities innew product projects are carried out in a qualityfashion. A complete and thorough process, where everyneeded activity is carried out without hasty comer-cutting. A flexible process, where stages and decision pointscan be skipped or combined, as dictated by the natureand risk oftheproject.2 . A clearly defined new product strategy for thebusiness unit, which m eans: There are goals or objectives for the husine ss's totalnew product effort (e.g., what sales, profits, etc. newproducts would contribute to the husiness). The role of new products in achieving husiness goalsis clearly comm unicated to all. There are clearly defmed areas of strategic focusstrategic arenasto give direction to the business'stotal new product effort.3 .Adequatenewproduct resources meaning: Com mitting the necessary people. Allowing them sufficient time. Providing an adequate R&D budget.

only one phase of project, or having project leadershipchanging hands many times during a project's life). All projects had an assigned team of players whoworked on specific projects. These assigned players, who executed projects, were across-functional teamfrom R&D, Marketing, Manu-facturing, Engineering, etc.Businesses that consistently used cross-functional teamswere rewarded: the total new product effort was moreprofitable on average. But the mere use of cross-functional teams did not have the dramatic impact onprofitability or impact that we had expected; rather, teamquality (No. 5 above) had the stronger effect.9. Senior managem ent accountability.This final driverhadtheleast effect overall on perfonnan ce, but it did helpto drive the profitability that the business unit achievedfrom new products. Managem ent accountability capturesthe degree to which new product performance wasmeasured, and senior management was held accountablefor the program results. Scores were very low on allingredients of management accountability. Considereach ingredient in greater d etail: New product performance w as a part of senior man-agement's person l performance objectives. Senior manag emen t'scompensationor bonus schemewas tied to the business uni t's new product performance .Very few businesses employed this practice.e Theperformance resultsofthe new product programwere actuallymeasured(e.g., percentage of annual salesgenerated by new products, or success, fail and killrates). Surprisingly, this practice was rarefirms simplyfailed to keep scoreThe evidence sugg ests that not only is the business u nit'snew product performance not a factor in senior manage-m ent 's comp ensation schem e, or even an integral facet ofsenior management's annual performance objectives,but such new product performance results are likely no teven measuredat all for a great many firms.Conclusions1 New product performance at the business-unit levelcan be reduced to two major underlying dimensionsprofitability andimpact on the business. This reductiongreatly simplifies the measu rement and reporting of per-formance. Note that there exist many measures of newproduct performancepercentage of sales, successrates, meeting sales and profit objectives, variousmeasures of profitability, and even technical successratings. We used ten of these many performance metricsin the current s tudy. One problem with mult iplemeasures is that the performance issue becomescloudedit is more complex and more confusing.

Research Technology Management


14/16

The current study provided needed clarification here.The many measures of performance used really boildown to two major dimensions, according to our statis-tical analysis results: How profitable the businesses's total new productefforts are, including: the profitability versus competi-tors and the overall success rating ofthe business's totalnew product effort; whether the total initiative meetsprofit objectives; its profitability relative to spending;and the impact ofthe total effort on the business unit'sprofits. The impact of the total new product effort on thebusiness, including: percentage sales by new productsachieved by the business unit; impact of new products onboth sales and profits of the unit; success rate achieved;and the technical success rating.Thus,when speaking o f t h e performance ofabusiness'snew product efforts," our research sug gests that it makessense to simplify the discussion and focus on tw o broadperformance areas:profitability andimpact (One couldeven reduce "new product performance" to two perfor-mance indices consisting of a number of metrics; theweights or loadings shown in Figure 2 enable the com-putation of these two indices). The Performance Mapshown in Figure proved to be a mo st useful visual toolto capture these two dim ensions and the multiple metricsthat comprise them. Indeed, it makes sense to plot indi-vidual business units within a corporation on this X-Ymap, in order to display pictorially the high-impactand/or the high-profitability business units.2.The three cornerstones of new product perform ancethe factors that really drive performanceare: A high-quality new product process. A clearly defined new product strategy for thebusiness unit. Adequate new product resourcespeople and money.These three together had by far the strongest effect onboth measures of the business's new product perfor-mance. Merely having a formal new product process inplace had no effect on performance. Rather, it was thenature ofthe proc essa high-qu ality process which builtin the ingredients listed in "The Cornerstones of NewProduct Performance (below )that made the d ifference.The message is clear. Technology or new productstrategy must be firmly linked to business strategy. Thismeans that management must develop a new productstrategy for the businessa product innovation charter(6)that ties new products closely to the achievement ofbusiness g oals, has clearly stated objectives, and definesareas of strategic focus or thrust.

May -June 2007

The third cornerstone of performance adequate newproduct resourcesis sometimes overlooked in thesedays of right-sizing and re-engineering. Managementmust realize that developing a grandiose strategy, orhir ing consultants to re-engineer the new productprocess, will not yield results if there are no or too fewplayers on the field. The goal of a high-performance newproduct effort and a high-quality new product processwill not be achieved unless the resources are in place.Management m ust make the commitment here, and keepit3 .The R& D spending devoted to new products was thesingle strongest driver of progra m impact This comes asno surprise, but it is reassuring to see the relationshipverified and quantified. If the business unit's goal is ahigh-impact new product effortone that yields a highpercentage of sales by new products, and has a largeimpact on the business 's sales and profi tsthenincreased new product R&D spending is one route. Putanother way, do n't expect to have a high-impa ct effort ifmanagement is not prepared to comm it the R&D dollarsas a percentage of sales. Note that the mean R&Dspending for the business units studied w as already rela-tively high at 7.5 percent ofsales.There are caveats here,however:o There was no relat ionship at al l between R&Dspending andtheprofitability ofthenew product effort.o The link between R&D spending and impact was notone-to-one (the correlation was strong at 0.395, but farless than1.00 .4.Other factors did enter the performance equation, butthese factors have far less effect than the three comer-stones of performance highlighted above. These lesserfactors are:


15/16

H igh-quality new product teams: Dedicated teamleaders, foil-time on their one project; good intem al teamcommunications; and the ability to manage the extemaldecision-making process. Senior management commitment: Senior managersstrongly committed to new products; commitment ofthenecessary resources; and senior management closelyinvolved in Go/Kill and spending decisions. Innovative climate and culture: A new product ideasubmission schem e; free time or scouting time av ailablefor technical em ployees; resources av ailable for creative,unapproved projects; and skunk works on unofficialprojects. Note that most firms were particularly weak onthis factor. Cross-functiona l teams: A defined and accountableleader who was accountable from beginning to end o ftheproject, and an assigned team of players from variousfunctions in the business. Senior management accountability:New product per-formance results were measured; new product perfor-mance was a part of senior man agemen t's performanceobjectives; and senior management's compensation orbonuses were tied to new product results. Businessesfared very poorly on these three accou ntability practices.5. Bew are naive process re-engineering Recall thatsimply having a formal new product processone thatmaps out the steps in a product development from ideato launch does not lead to positive performance,according to our s tudy 's resul ts . Many f i rms arecurrently busy documenting their product innovationprocesses, simply to m eet ISO standa rds; the end result isa "formal process" but likely no increase in perfor-mance If you naively reengineer your innovationprocessthat is, simply document what you are nowdoingdon't expect positive results. Rather, considerthe success factors of a high -qua l i ty innov at ionprocessthose factors listed in conclusion 2 aboveand build these into your new product process. Manyfirms have utilized the stage-gate model as a basis forredesigning their new produc t process (7); the stage-gateapproach emphasizes many of the ingredients listedabove: homework, voice of the customer, tough gates,quality of execution; and fiexibility. Some progressivefirms have even moved beyond this into third-generationnew product processes (5).

6. Benchmark before re-engineering. The beginning of asolution is understanding the problem. Before youembark on a re-engineering "solution," take time tobenchmark your business uni t ' s performance andpractices against other firms. The ten performancemetrics and the 48 measures of practices used in ourresearch are a good starting point (i.e., become thelogical variables or measures to use in such a benchm ark-ing exercise).In subsequent work, we have extended our analysis anddeveloped our ProBE (for Product Benchmarking andEvaluation) benchmarking tool, which enables you tocompare your new product performance and practicesagainst the many business units in our database. Ifyou cannot afford extensive benchmarking, then atminimum, perform a diagnostic on your own businessunit; that is, rate your own performance and practices onthe ten metrics and48measures of new product practiceswe have used.In sum, new product developm ent, mo re than ever, is tiedto corporate fortunes and prosperity. An understandingofthefactors that drive new produ ct performance at thebusiness unit level is critical if we are to achieve the goalof increased performance. By identifying these drivingfactorsin particular, a high-quality innovation proce ss,a defined innovation strategy, and adequa te resources tosupport both strategy and processthis benchmarkingstudy has provided hard and quantified evidence formanagement to take the steps toward winning at newproducts. References1. Cooper, R. G. "Debunking the Myths of New Product Develop-ment."/Jeiearc/i TechnologyManagement2:1,A {ivi\y A\igwsX 19940-50.2. Montoya-W eiss, M. M. and R. J. Calantone. "Determ inants ofNew Product Performance: a Review and Meta Analysis." Joi/rna/oProduct Innovation Management 11,5 (Nov. 1994: 397-417.3. Cooper, R. G. and E. J. Kleinschmidt. "Benchmarking the Fin n'sCritical Success Factors in New Product Development." Journal oProduct Innovation Management12, 5 (Nov. 1995):374-391.4. Cooper, R. G. and E. J. Kleinschmidt. "Benchm arking Firm sNew Product Performance and Practices."Engineering M anagemenReview 23,3 (Fall 1995): 112-120.5. Product Launch Road Map for Success.Procter & Gamble. Cincinnati: P&G internal document (July 1993).6. Crawford, C. M. "Defming the Charter for Product Innovation."Sloan Management Review (1980): 3-12.7. Cooper, R. G.Winning at New Products: AcceleratingtheProcesfrom Idea to Launch.Reading, M ass: Addison-Wesley, 1993.8. Cooper, R. G., "Third-generation New Product Proces ses."Journal of Product Innovation M anagement 11 , (Jan. 1994): 3-14

Subscribers Read RTM OnLine . . .. . . as soon as the issue is printed, at http://www.iriinc.org/rtm.Full text, electronic version is searchablewithin text and across other joumals.

I cehnology Managemeiif


16/16

winning businesses in product development

Documents