Innovation and Re-Innovation

l of MarktMg Maiutgenrnt, 1986, 2, No. 2, 10<t-123

Innovation and Re-Innovation: A Role for the User

Roy RothwellLeader, Design and Innovation Management Group, Science Policy Research Unit, University of Sussex

Various models of the innovation process, from the "traditional" to the more recent, areexamined in this paper which focuses on the implications for the role of the user in theinnovation process. The author points out the need for greater recognition of theimportance of users as active participants in the innovation process. In many industrialsectors user-need specification and product development involve more than simply apassive role for the user and innovatory success is associated with active user involvementin product specification, design and development.

INTRODUCTION

All modern studies of the technological innovation process emphasise theimportance of demand specification to innovatory success. In other words,and not surprisingly, it is necessary for product characteristics to match the"user needs" profile (Rothwell, 1977a; Cooper, 1980). This implies the needfor a strong marketing effort on the part of would-be innovative firms; theneed to scan the marketplace in order to establish an optimum andrepresentative set of user need specifications as the basis for the new productdesign brief

This market scanning process, the traditional task of the marketer, in asense implies a passive, or at best a reactive, role for the user; he simplyresponds to the questions of the marketer who then goes away andestablishes the appropriate design brief on his behalf In some market areas,for example, in the consumer products area, users might very well play asimple reactive role. In other areas, however, for example industrial goods,the user might play a much more active role, not just in user needspecification, but also in the technological development process itself Someevidence for the importance of users as active participants in the innovationprocess will be briefly presented below.

MODELS OF THE INNOVATION PROCESS

We will begin by describing a number of "traditional", and several morerecent models of the innovation process, and in particular their implicationsfor the role of the user in innovation.

109

110 ROY ROTHWELL

(i) Technology-pash

Early models of the innovation process emphasised the causal roie ofsciendfic and technoiogical advance and were generally iinear. These can besummarised by Figure 1 (a) which is the so caiied "technoiogy-push" modeiof innovation. According to this modei, discoveries in basic science leadeventuaiiy to industriai technologicai deveiopments which iead in turn to aflow of new products and processes on to the marketpiace. This modeiimpiies a more or iess passive roie for the user, and the marketpiace issimply a receptacle for the resuits of firms' scientific and technoiogicaiendeavours.

To be successfui on the basis of the modei, the firm requires mainiy tokeep in contact with deveiopments in basic science and to maintain asubstantial technoiogicai deveiopment capabiiity. The basic premise under-iying the modei is that "more R&D" is equivalent to "more innovation".Whiie some notabie deveiopments have occurred following the sequencedescribed in Figure 1, e.g. X-ray photography in medicine, in generai it ismore iikeiy to result in the production of "better mousetraps'. Even todaypubiic policies towards technological deveiopment appear, impiicitiy atieast, to foiiow the technoiogy-push model: they emphasise suppiy-sidefactors (R&D) and iargeiy ignore the roie of demand (market) factors.

(a) SCIENCE DISCOVERS, TECHNOLOGY PRODUCES, FIRM MARKETS

BasicSctence

AppliedScience andEngtneeiing

Manufactifting Marketing

(b) NEED PULLS, TECHNOLOGY MAKES, FIRM MARKETS

MarketNeed

Development Manufacturing Sales

Source: Rothwell (1983).

FIGURE 1 Two extreme models of the innovation process—the "traditional" views.

(ii) Need-puU

From the mid-1960s onwards, iargeiy as the resuit ofa growing number ofempiricai studies and descripdons of actuai innovations, the roie of feit needbegan to be emphasised. One of the most wideiy quoted statistics in theinnovation iiterature during that period was that "75 percent of successfulinnovations arise in response to the recognition of a market need" (SeeTabie 1). This was interpreted into the iinear "need-puii" modei ofinnovation shown in Figure I(b). According to this modei, innovations arise

INNOVATION AND RE-INNOVATION 111

1 ABLE 1

Summary of fmdings on source of stimulation for innovation in eight empiricalstudies.

Study

Baker et al.Carter and WilliamsGoldharSherwin and IsensonLangrish et al.Myers and MarquisTannenbaum el at.Utterback

Proportion of innovationsfrom market, mission or

production needs ("„)

7773696666789075

Proportion fromtechnical

opportunities {%)

2327313434221025

Source: Voss, C. A., 1984.

as the result of perceived and often clearly articulated needs, a goodillustrativ e example being radar.

The adoption of the need-pull model led to the question being asked insome firms, "who needs R&D?" and in particular "who needs research?"After all, if ideas for innovations flow freely from the marketplace, thenperhaps only a limited development capability is necessary to transformthem into new products. This, of course, was an extremely short-sighted andmisguided view since, when major market changes occurred, or when.significant new technological capabilities emerged, the firm could find itselfin the position of being unable to respond. Firms adopting this approachran the risk of being Iocked-in to a regime of incremental innovation inexisting areas as a result of innovating only to meet gradually changing userspecifications. In addition, this model generally assumes a traditional"reactive" role for the user, implying simply the need to scan themarketplace in order to establish the modified design specification.

(iii) The coupling model

During the past decade or so, both the pure technology-push and need-pullmodels of innovation have increasingly been regarded as extreme anduntypical examples of a more genera! process of coupling between science,technology and the marketplace (Mowery and Rosenberg, 1978). It is nowgenerally accepted that more R&D has not necessarily resulted in moreinnovation and, as Hayes and Abernathy (1980) have demonstrated,overemphasis on current market needs has indeed led in some cases to aregime of incrementalism and lack of radical innovations. In fact, in theearly inventive phases of a radical innovation, users may have only a

112 ROY ROTHWELL

glimmering of the eventual potendai of the new deveiopment. "The WrightBrothers' proving of the feasibiiity of powered manned flight, for exampie,did not and couid not sdmulate even the most visionary of onlookers to seewhat eventuai user needs wouid be iike for a commercial mail or passengerairiine service" (Rothweli and Gardiner, 1985).

A more representative modei of the innovation process in practice is givenin Figure 2. This is the so-caiied "coupiing modei". According to thismodei, innovation can be regarded as a iogicaliy sequentiai, though notnecessarily continuous process, that can be subdivided into a series offunctionally separate but interacting and interdependent stages. The overaiipattern of the innovation process can be thought of as a compiex net ofcommunication paths, both intra-organisadonal and extra-organisational,linking together the various in-house functions and linking the firm to thebroader scientific and technicai community and to the marketplace. In otherwords, the process of innovadon represents the confiuence of technologicalcapabiiities and market needs within the framework of the innovadng firm.

New need

iIdea

conception

Itechnological

capability

Needs of societv and the market place

1Deveiopment

1

\

1 1 'Manufacturing Marketing

and saies

1 1

\

fMarket

1State-of-the-art rn technology and production techniijues

Source: Rothwell and Zegveld (1985).

FIGURE 2 Interactive model of the innovation process.

The coupiing modei suggests that it is not particuiariy cruciai whetherthe new product concept arises as the resuit of the emergence of a newtechnoiogicai capabiiity or because ofthe recognition ofa new market need(or, as is most iikeiy the case in pracdce, a combinadon of both). Theimportant factor is that the firm coupies technoiogicai capabiiity to marketneeds at the eariiest possibie stage. Moreover, as the project deveiops, thefirm remains coupled to the marketpiace in order that changing marketrequirements can be detected and fed in to the development process toproduce a modified design brief. Whiie in this model the user has someinvolvement throughout the course of the deveiopment in specifying hischanging requirements to the innovator, a technicaiiy active user roie is notnecessarily implied.


(iv) Re-innovation

The above models imply that innovation is in some sense a clearly boundedprocess that terminates once the original new product reaches the market-place. In practice, technological innovation is a dynamic, iterative processrather than a one-off event. A successful new design represents only atemporary balance between product characteristics and user requirements;continued success requires frequent adjustments to this balance. Innovationas an iterative design process is illustrated in Figure 3, which shows theprocess of re-innovation to meet altered user need specifications.

Most marketers would accept that users can play an important role in re-innovation. They feed back information to the manufacturer concerningchanging conditions of use, greater ease of use, improved serviceability andso on. In general, however, the bulk of real technological change isgenerally assumed to continue to arise as the result of the manufacturers'own R&D activities.

Bnicidn

concept

Designfor

demonstration

Mkt

Inventton

Practical and materialembodiment of thebasic idea or concept

Design for maritetabtepTDduction

Innovation

Commercial ej^ioitationof the basic idea orconcept i e. specificproduct characteristicsat 8 particutar price

Key

L Re-innovaiion

Altered productartd proemsspecificationse.g. impfovedperformance atlower cost

Patents, registered designs and trade marks

taken out fof jvotectron

Source: Rothwell and Gardiner (1983).

FIGURE 3 Invention, innovation and re-in nova tion.

USER-PRODUCER LINKS AND SUCCESSFUL INNOVATION

In this section the results of a number of empirical studies will briefly bereviewed that have illustrated the importance of user-producer linkages tosuccessful innovation. The first of these, a European-wide study that was

114 ROY ROTHWELL

undertaken in the mid-1970s (Rothweii, 1976a), deals with a traditiona!sector of industry, textiie machinery.

Textile machinery is a sector in which British international competitive-ness deciined markedly from the iate i 950s onwards. One of the reasons forthis decline was a lack of embodied technoiogicai change in Britishmanufactured machinery. Moreover, even where firms had been innovadve,their innovations often faiied satisfactoriiy to match user requirements. Thispoint is underlined in the data presented in Figure 4.

ImportantI

Verv importantI

General charactenslics

Cheaper overall running costs

— 1 5

Cheaper overall instaNatton costs

—I 7

— —I B

Specific characteristics

Productrorr rate

Improves product quality

Reduced labour requirements

1 7

Purchase pirce

— H14

Reduce no of product operations

9

Key

Teictile machrnerv manufacturers

—— Textile machrnery users

FIGURE 4

The data in Figure 4 are taken from a comprehensive questionnairesurvey of UK textiie companies (users) and UK textiie machinery manu-facturers (producers) (Rothweii, i977b), in which respondents were re-quested to weight a iist of eighteen specific machinery characteristics andnine rather more generai characteristics according to their degree ofimportance, on a five point scale ranging from "not important" to "ofcrucial importance". From these data the rank order of importance wascomputed for each of the characteristics.

The bars on Figure 4 indicate the weighting or degree of importanceattached to two of the "general" characteristics and five of the "specific"


characteristics by both users and producers, and the numbers indicate theirrank order of importance. It can be seen, in a number of instances, thatthere is a notable mismatch in perception of importance weighting, rankorder of importance, or both. For example, while users and producers gavethe factor "production rate" equal ranking, users weighted it more highlythan producers; while both gave equal weighting to the factor "purchaseprice", users ranked it well below producers.

Perhaps the most significant difference occurred with the general charac-teristic "cheaper overall running costs", which was awarded a much higherweighting by users. This, in effect, is a composite of a number of specificfactors including efficiency (reliability), production rate, labour require-ment, purchase price and working life. Essentially, users were very muchmore aware of the importance of total life cycle costs than producers whowere more concerned with one component of this—purchase price. Thepoint is, this, and other mismatches in perception could have been avoidedhad producer companies established active collaboration with the userduring machinery' design and developtnent. Unfortunately, all too fewBritish companies involved the user in the design/developtnent process, userinvolvement generally not occurring in any substantive sense until just priorto, or following, commercial launch.

Not all British textile tnachiner)' companies were remiss in this respect,howev er, and this is illustrated in Table 2 which shows patterns of externalcollaboration during the production of 25 successful UK textile machineryinnovations. Of the 25 companies involved, 85 percent enjoyed externalcollaboration during development, of which nearly 70 per cent was withpotential customers. Moreover, a number of companies interacted acrossseveral stages of development and with several outside agencies. In all cases.

TABLE 2

Patterns of outside collaboration during the production of 25 successfulinnovations introduced bv the UK textile machinerv' industr\\

Outside collaborationNo outside collaborationStage at which collaboration sought R&DPrototype productionProductionMarketing and sales

Collaboration was with*

CustomerOther industryOther

, \ umber

214

131361

1452

Percentage

8416

> 1

>4~ ^ I-> 1

662410

•Percentages calculated to nearest integer.

116 ROY ROTHWELL

especiaiiy those invoiving the user, the interacdon involved much more thansimpiy consuitation; in each case the user was an active partner in the actuaidevelopment of the machine.

The importance of active customer involvement in machinery deveiop-ment was much more wideiy acknowiedged in the considerabiy moresuccessfui West German and Swiss textiie industries, whose constituentmanufacturers had strong tradidons of invoiving users in their machinerydesign procedures. A striking exampie of this was the enormously successfuiSulzer weaving machine—the worid's first shuttleless loom—that passedthrough ten prototype stages over many years before the commerciaiiaunch of the Mark 11 version. Throughout its deveiopment, the Suizerioom was tested regularly in the weaving mills of potential customers and itwas feedback from these that provided much of the impetus for continuedredesign (Rothweli, 1976b).

Further evidence supporting the importance of user-producer iinks comesfrom a comparadve study of the UK and West German machine toolindustries by Parkinson (1982). Parkinson found a considerabiy greaterpropensity on the part of the generaiiy much more successfui West Germanmachine tooi suppliers activeiy to invoive the user in the product design anddeveiopment process:

"In the West German companies in the sampie, customer invoivementin the product design and deveiopment process was seen as axiomatic ifthe company wanted to be successfui. In contrast, in the Britishsupplier companies, the prevailing attitude was not to involve thecustomer in the process untii the product was put on the market".

Effectiveiy, whiie British users and suppiiers swapped commerciai data(price, deiivery), the West German users and suppiiers swapped technicaland "user needs" data. A similar difference was evident between the Britishand West German textile machinery industries in their relationships withcustomers in the textiie sector.

Whiie it is important, where feasible, activeiy to invoive customers in thedesign and deveiopment process it is, of course, important to interact withcustomers who themseives are innovative and technicaiiy progressive. Wherethe home user industry is generaiiy technicaiiy unprogressive or technicaiiynon-demanding, this can present the manufacturer with considerabie prob-iems. Such, in fact, was the case with the UK textiie industry during the1950s and eariy i960s, much of which showed a marked reiuctance to adoptinnovative new machinery. This placed the UK textile machinery industryat a marked disadvantage in reiation to its major West German competitorswho enjoyed an innovation-demanding home market.

Further evidence supfKirting the importance to suppiiers of the existenceof innovadve users comes from Parkinson's study cited above:

"West German customers tended to vaiue technicai attributes of themachine tooi (inciuding sophisdcation) more than British customers,

INNOVATION AND RE-INNOVATION 11 7

and were more likely to suggest ideas for product modification orentirely new products to suppliers... Given that West German cus-tomers appear to have a more positive attitude to new technology thantheir British counterparts, it is clear that their role may be crucial inthe new product development efforts of their own domestic suppliers.The comparative weakness of British customers appears to go some waytowards explaining the comparative weakness of their domesticsuppliers. (An active user may create an active supplier.)"

Essentially, what is required for successful innovation in the capital goodssector is the establishment of a virtuous circle involving active collaborationbetween innovative suppliers and innovation-demanding users, each partnercontinually feeding performance information back to the other. Moreover,this process should begin as early as possible in the product design anddevelopment process.

User-producer interactions during innovation do not, of course, alwaysprogress smoothly and harmoniously and, in the case of major andexpensive innovations, both partners can take considerable financial risks.Such was the case with the development of the Boeing 747, which involvedintensive interaction between Boeing (the producer) and Pan Am (the user).When Boeing and Pan Am jointly signed an agreement in December 1965to develop and to operate respectively the 747, Boeing was comtnitted to aS2,000 million development programme which would be lost if the 747 wasnot a success, while Pan Am was committed to staged payments of $500million for 25 aeroplanes, with 50% at least six months before takingdelivery of a single 747. Between signing of the original agreement anddelivery of the first 747, the specification changed considerably, e.g.capacity from 350 seats to 374 seats and gross take-off weight from 550,000lbs to 710,000 lbs. Gardiner and Rothwell (1985) described user/producerinteractions during this period as follows:

"Phase II in our robust design model is a period of design consoli-dation which, for Boeing and Pan Am, turned out to be a four-year-long'nightmare' rather than one of progressive and cooperative develop-ment. The central problem was that as design detailing began theproposed take-off weight for the 747 kept progressively increasing. Inturn, this resulted in reduced cruising speeds, shorter ranges andreduced operating altitudes. During 1966 the initial cruising altitudedropped from an earlier 35,000 feet, which put it above the thenexisting 707s and DC-8s to just 33,000 feet, which put it back amongthe more congested airlanes occupied by the 707s and DC-8s. By 1968the take-off weight continued to increase and, when Pan Am began towork out operating cost per seat-mile, the proposed 747 on longerroutes would have been more expensive to fiy than the already existinglong range versions of the 707 Boeing, and Pan Am were by then intoa prolonged period of contractual dispute involving a great variety ofproposals and counter-proposals. At one time or another there were up

i i8 ROY ROTHWELL

to seven different versions proposed of which oniy the 747-iOOB and747-1 OOF were ever to emerge, and even then they had performancespecifications somewhat removed from those proposed during 1967 and1968". (Gardiner and Rothweii, 1985).

Whiie the sometimes fraught relationship between Boeing and Pan Amenjoyed a positive outcome, i.e. it resuited in the production of a highiysuccessfui pas^senger aircraft, in some cases potentiai users can piay anegative roie through resisting or retarding the introduction of innovations,especiaiiy radical innovations. At the same time, as pointed out earlier, itmight not be possible to even begin sensibly to specify user requirementsduring the eariy deveiopment stages of a radicai innovation. In such cases,user invoivement emerges during the latter stages through, perhaps, ex-perience gained in prototype tests or even in commerciai operadon. Thesefactors to some extent figured during the deveiopment of the Hovercraft.

"When Cockereli first began demonstrating his peripherai jet hover-craft effect, there do not seem to have been any immediate or specificusers in mind.. . Mainiy this was a period of non-customer directedexperimentation. Because user needs and specifications were diffuse andweakiy perceived, the design briefs and demonstrations tended to berather generaiised in nature. As the annular jet patent appiicadon wasbeing deveioped in the mid-1950s, the first potentiai user, in the formof the British military, rather than being supportive at this eariy stagein the deveiopment of a new technoiogy had, if anything, a rathernegative infiuence". (Rothweii and Gardiner, 1985).

Foiiowing a series of demonstrations proving the principie of hover flight,and as the Hovercraft deveioped, the roie of user specifications wasincreasingly important in shaping its design and performance targets:

"Having surmounted the initiai hurdie of demonstrating the commer-ciai practicaiity of its radiai technoiogicai design, the roie of actuai(and potentiai) hovercraft users became increasingiy important insubsequent deveiopments. The eariy, rather generaiised, design concepthad to be made more specific to meet the requirements and specifi-cations of a particuiar user or class of users. Once some producer/userexperience had accumuiated, this detaiied specification became eventighter at the re-design stage when the design brief contained a numberof precise "user specific" targets, e.g. speed, capacity, capitai costs andoperating costs with, in addition, weii understood reiationships betweenthese target figures". (Rothweii and Gardiner, i985).

USERS AS INNOVATORS

It was, without doubt, the work of E. von Hippei (1976, 1978, 1979a)which first demonstrated, in any systematic sense, that in certain sectors of


industry users can play the major role in invention and early innovation.Von Hippei's pioneering work covered four families of scientific instruments:gas chromotography, nuclear magnetic resonance spectrometry, ultravioletabsorption spectrometry and transmission electronic microscopy. In eachcase, the instrument was first developed by a user and later transferred to amanufacturing company for commercial production. In addition, not onlydid the user play the dominant role in invention, but he also played anactive part in re-innovation. Thus, from a total of 44 major improvementsto the instruments, customers were the source of 36 (80%) and, from a totalof 46 minor improvements, 34 (73%) derived from users.

On the basis of his research results, von Hippel postulated a customer-active paradigm (CAP) and a manufacturer-active paradigm (MAP) forindustrial product idea generation; in the former the customer plays the keyproduct-initiating role, while in the latter he plays merely a reactive role(von Hippei, 1979b).

In attempting to explain his results, von Hippel utilised the concept ofthe appropriabiiity of innovation benefit as a predictor of the functionallocus of innovation. Commenting, for example, on the work of Berger(1975) and Boyden (1976) who found, in the plastics and plastics additivesindustries respectively, that all the innovations studied derived exclusivelyfrom the manufacturer, he states:

"As noted earlier Berger and Boyden have, for example, sampledplastics and plastics additives respectively and have found all of theseto have been developed by product manufacturers rather than byproduct users. I suspect that further research would show this locusexplicable in terms of the ability of users and manufacturers toappropriate output-embodied benefits from these categories of inno-vations. A particular plastic or additive is typically not essential tousers since other materials exist which can do the job at a (usuallyminor) cost premium. To the manufacturer, however, a plastics andadditives innovation which provides such a slight cost advantage mightmean that users of other materials (steel, aluminium, other plastics,etc.) replace these with the innovative material and quickly becomemajor customers, thus allowing the innovator to capture significantoutput-embodied benefit". (Von Hippel, 1979a).

While the appropriabifity of benefit undoubtedly is a crucial factor indetermining the functional locus of innovative activity, a second factorshould be added, that of the locus of state-of-the-art expertise. It is not, forexample, surprising that medical researchers are the source of a significantpercentage of medical instrument innovations (see below) since they areinvolved in state-of-the-art medical research which often necessitates themconstructing new kinds of monitoring and measuring equipment (in col-laboration, in the UK, with medical physicists and technicians). The sameis true of chemists in government and university laboratories, who oftenneed to invent and build new analytical equipment to further their

120 ROY ROTHWELL

pioneering research. Medical researchers are not, of course, normaily in thebusiness of instrument manufacture, and they can best appropriate thebenefits of their inventive acdvity through buying back professionaiiymanufactured, easy to use and reiiabie instruments from manufacturers towhom they have transferred their invendon. Licence income can aiso beused to fund further research. The manufacturer, on the other hand,appropriates the benefits of his design, engineering and manufacturingefforts through saies receipts and increased market share. It wouid beunusuai for the typicai instrument manufacturer to contain the ievel ofmedical research skiiis necessary to enable him to perceive the need for, andto make, the initial invention, although he can remain "plugged-in" to theinventive, state-of-the-art medical researcher and udiise him as a continuingresource in the process of re-innovadon.

In a recent articie in this Journai, Foxail el al. (1985) have extended vonHippei's concept of a spectrum of user invoivement in product-initiationranging from the MAP (iow) to the CAP (dominant). The authorsproposed a CAP2 model in which the user piays an entrepreneurial roie inthe deveiopment of user-initiated products; they then iiiustrated thisthrough describing the business development activides of British Aerospace'sWarton Division. According to CAP2, the user piays an acdve part not oniyin invention, but aiso in development of his invention for commerciai expioit-ation, thus appropriadng internaiiy the benefits that wouid otherwise beobtained by the manufacturer as in von Hippei's original CAP modei.

In a recent series of articies Shaw (1985, 1986a, 1986b) has described theresuits of a detaiied research project into user-producer interactions in theUK medical equipment industry. He found, from a sampie of 34 inno-vations, that 26 (76%) were deveioped through "muitipie and continuous"user-manufacturer interaction, resulting in 22 (65%) being successfui, onetoo eariy to judge and three being faiiures. Moreover, aii the faiiures weredue to the unsadsfactory technicai performance of the equipment, user-needs being appropriately specified in each case. According to Shaw, "this79% success rate far exceeds the 55 to 64% success rates found byresearchers in other industries and the degree of transfer of sociai benefit toprivate innovadon benefit is much higher".

In Shaw's sampie of medicai equipment innovadon, user-invoivement inmost cases continued after the transfer of the initiai device to the manu-facturer. This inciuded joint prototype deveiopment and testing, user-evaiuation and assistance with marketing. In a number of cases, presdgiousmedicai researchers (inventive users) often provided favourable pubiicity forthe innovations by presendng the resuits of research (using the innovadons)to national and intemationai meetings. In this respect, they were seen as"honest brokers" by their peers, who were aii potential customers.

DISCUSSION

The innovadon iiterature strongly underlines the fact that a significant


percentage of unsuccessful innovations fail because the innovator has notsucceeded in satisfactorily establishing an appropriate set of user specificationsand in interpreting these in the design of his new equipment. In manyindustrial sectors, user-need specification and product development involvemore than sitnply a passive role for the user, and innovatory success isassociated with active user involvement in product specification, design anddevelopment. Uset^ also have an important role to play in the process of re-innovation. Moreover, it is users who are themselves technically-progressiveand innovation-demanding who have the greatest potential in this respect.

In some product areas, empirical evidence points to an inventive role forthe user, who in this case is the primary source of the basic technologicalknow-how. In areas characterised by user-invention, there are obviousadvantages to both partners. On the side of the manufacturer, for example:user needs are effectively built-in to the invention; R&D is effectivelysubsidised; the user plays a marketing role through the dissemination ofresults obtained using the equipment; and the user plays an important rolein demonstrating the effectiveness of the equipment which, because of hisactive role in its development, he uses optimally right from the start. Onthe side of the user: he obtains well engineered and reliable equipment builtto satisfy his requirements; he gains royalties on sales to fund furtherresearch/development; and first use of the equipment enables him to gain aresearch lead over his peers.

The potential ability of the user to play an important role in the re-innovation process has important implications for product design. Specifi-cally, it would seem sensible for manufacturers to design products that areamenable to adaptation and improvement by innovative users, and thispoint was illustrated in an important piece of research by von Hippel andFinkelstein (1979):

"We therefore decided to explore our speculation that different types offunctionally similar products (here, different types of automated clinicalchemistry analyser) could show significant differences in the innovationrole of the user by comparing extant user test development activityinvolving different types of automated clinical analyser".

Von Hippel and Finkelstein compared, in detail, two chemical analysers;the Du Pont ACA and the Technicon SMAC. They found that of 18 newchemical test methods offered commercially by Du Pont, all had beendeveloped in-house. In the case ofthe Technicon analyser, however, 74% of20 new test methods had been developed by users. The principal reason forthis difference lay in certain design features of the Du Pont equipmentwhich militated against user-test-method developments, notably the designof its reagent system, which was much less amenable than its Techniconcounterpart to user-adaptation. In terms of hardware-embodied innovations(equipment improvements), the single equipment improvement identified inthe Du Pont analyser (as well as the original innovation itself) wasproduced in-house. In the case of the Technicon analyser, the original

122 ROY ROTHWELL

innovation and 46% of important improvement innovadons were made bythe user.

The lesson is ciear. Not oniy shouid wouid-be innovators acdveiy invoivethe user in product specification, design and deveiopment; not oniy shouidthey seek in pardcular to invoive technicaiiy progressive users; but theyshould also offer product designs that are user-fiexible. Manufacturersignore this message at their peril.

References

Berger, R. (1975), "Factors influencing the locus of innovation activity leading to scientificinstrument and plastics innovation", Unpuhlished SM Thesis, Sloan School of Manage-ment, MIT, Cambridge, June.

Boyden, J. (1976], "A study of the innovation proces.s in the plastics additives industry",Unpublished MS Thesis, Sloan School of Management, Cambridge, MIT, January.

Cooper, R. G. (1980), Project Newprod: What Makes a New Product a Winner, QuebecIndustrial Innovation Centre, Montreal, Canada.

Foxall, G. R., Mutphy, F. and Tierney, J. 11985), "Market development in practice: A casestudy of user-initiated product innovation". Journal of Marketing Management, 1(2), Winter,pp. 213-221.

Gardiner, P. and Rothwell, R. (1985), "Tough customers: Good designs". Design Studies,6(1), January.

Hayes, R. H. and Abernathy, VV. J. (1980), "Managing our way to economic decline".Harvard Business Smea, July-.^ugust, pp. 67-77.

Mowery, P. C. and Rosenberg, N. (1978), "The influence of market demand uponinnovation: A critical review of some recent empirical studies", Research Policy, 8, April.

Parkinson, S. (1982), "Successful new product development: Having a good customer helps".The Business Graduate, Special Issue, Innovation, 12(!j.

RothweU, R. (1976a), innovation in Textile Mackintry: Some Significant Factors in Success andFailure, Science Policy Research Unit, Occasional Paper Series No. 2, University of Sussex.

Rothwell, R. (1976bi, "The Sulzer weaving machine: A case study of successful innovation".Textile Institute and Industry, May, pp. 170-173.

Rothwell, R. (1977a), "The charactenstics of successful innovators and technically pro-gressive firms", R&'D Management, 7(3), June.

Rothwell, R. (1977b), "Users' and producers' perceptions of the relative importance ofvarious textile machinery characteristics". Textile Institute and Industry, July, pp. 239-242.

Rothwell, R. and Gardiner, P. (1983), "The role of design in product and process change".Design Studies, i{3).

Rothwell, R. and Gardiner, P. (1985), "Invention, innovation, re-innovation and the role ofthe user: A case study of British hovercraft development", Technovation, 3, pp. 167-186.

Rothwell, R. and Zegveld, "W. (1985), Reindustrialization and Technology, Longman, London.Shaw, B. (1985), "The role of the interaction between the user and the manufacturer in

medical equipment innovation", R&D .Management, 15(14).Shaw, B. (1986a}, "Appropriation and transfer of innovation benefit in the UK medicai

equipment industry", Technovation, 4, pp. 45-65.Shaw, B. (1986b), " I h e roie ofthe interaction between the manufacturer and the user in the

technological innovation process, PhD Thesis, Science Policy Research Unit, University ofSussex.

V'on Hippel, E. (1976), "The dominant role of users in the scientific instruments innovationprocess", Research Policy, 5(3).

Von Hippel, E. (1978), "Usere as innovators". Technology Review, 80(30), January."Von Hippel, E. (1979a), "A customer-active paradigm for industrial product idea gener-

ation". In: Industrial Innovation. (Ed.) Baker, M.J . Macmillan, London.


Von Hippel, E. (1979b), "Appropriability of innovation benefit as a predictor of thefunctional locus of innovation", Sloan School of Management, M i l , Working paper 1084-79, June.

Von Hippel, E. and Finkelstein, S. (1978), "Product designs which encourage—ordiscourage—related innovation by users: An analysis of innovation in automated clinicalchemistry analysers", Working paper 1011-78, Sloan School of Management, Cambridge,MIT, July.

Voss. C. A. (1984), "lechnology push and need pull: A new perspective", R&D Management,14(3).

Innovation and Re-Innovation

Documents