Eight types of product–service system: eight ways to sustainability? Experiences from SusProNet

EIGHT TYPES OF PRODUCT–SERVICE SYSTEM: EIGHT WAYS TO SUSTAINABILITY? EXPERIENCES FROM SUSPRONET

Arnold Tukker*

TNO Institute of Strategy, Technology and Policy, The Netherlands

Copyright © 2004 John Wiley & Sons, Ltd and ERP Environment

Function-oriented business models orproduct–service systems (PSSs) are oftenseen as an excellent means for achieving‘factor 4’. SusProNet, an EU network onPSSs, showed a more complicated reality.At least eight different types of PSS exist,with quite diverging economic andenvironmental characteristics. Theeconomic potential of each type wasevaluated in terms of (i) tangible andintangible value for the user, (ii) tangiblecosts and risk premium for the provider,(iii) capital/investment needs and (iv)issues such as the providers’ position inthe value chain and client relations. Theenvironmental potential was evaluated bychecking the relevance of certain impactreduction mechanisms (e.g. moreintensive use of capital goods, inherentincentives for sustainable user andprovider behaviour etc.). Most PSS typeswill result in marginal environmental

Business Strategy and the EnvironmentBus. Strat. Env. 13, 246–260 (2004)Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/bse.414

improvements at best. The exception isthe PSS type known as functional results,but here liability and risk premiumissues, amongst others, need a solution.Copyright © 2004 John Wiley & Sons, Ltdand ERP Environment.

Received 4 March 2004Revised 22 March 2004Accepted 14 April 2004

INTRODUCTION

Aproduct–service system (PSS) can bedefined as consisting of ‘tangible prod-ucts and intangible services designed

and combined so that they jointly are capableof fulfilling specific customer needs’ (see e.g.Tischner et al., 2002).1 Many see PSSs as anexcellent vehicle to enhance competitivenessand to foster sustainability simultaneously. Inmany markets products have become equallywell performing commodities. Here, the basisof competition has shifted from functional per-* Correspondence to: Arnold Tukker, Programme Manager Sus-

tainable Innovation, TNO Institute of Strategy, Technology andPolicy, P.O. Box 6030, 2600 JA Delft, The NetherlandsE-mail: [email protected] 1 Though the term PSS is now fairly well know in circles of (sus-

tainability) researchers and also widely used in the EU’s FifthFramework Research Program (FP5), business usually use termssuch as ‘solution development’ or ‘functional sales’.

EIGHT TYPES OF PRODUCT–SERVICE SYSTEM

Copyright © 2004 John Wiley & Sons, Ltd and ERP Environment Bus. Strat. Env. 13, 246–260 (2004)

247

formance, via fitness for use, to price (see e.g.Montalvo Corral, 2002, p. 20). A PSS businessmodel allows firms to create new sources of added value and competitiveness, since they

• fulfil client needs in an integrated and cus-tomized way, hence allowing clients to concentrate on core activities,2

• can build unique relationships with clients,enhancing customer loyalty, and

• can probably innovate faster since theyfollow their client needs better.

In theory, the ‘result-oriented’ PSS has the par-ticular potential to overcome all kinds of splitincentive concerning the environment (UNEP,2002, p. 9). As a result, it would be in the interest of both producer and consumer tominimize life-cycle costs, and hence the use ofconsumables in the use phase.3 Furthermore, ifone tries to design a need fulfilment systemthat really takes final consumer needs (ratherthan the product fulfilling the need) as a start-ing point, the freedom to design systems withfactor 4–10 sustainability improvements ismuch higher.

These benefits prompted the EU to investheavily in PSS development in its Fifth Framework Programme, partly by setting upSusProNet (the acronym for the SustainableProduct Development Network).4 However,SusProNet’s analyses thus far do not confirmthat PSSs are a win–win strategy. For manysuccessful PSSs it is not clear whether they are sustainable, and many purpose-designed

sustainable PSSs (pay per wash, sustainablegrocery conscriptions etc.) have failed, or clingon to a marginal existence in the market.5 Sowhy is this? In our view, analyses of PSSs oftenforget to focus on the following points.

• A major reason for being in business is tocreate added value. Industry’s new businessdevelopers will embark on PSS models,when that is the key to competitiveness andvalue creation, but will adopt other modelswhen that better helps to beat competition.The first question to be answered in thispaper is which factors determine whether a PSSbusiness model is the best way to create valueadded?

• Sustainability is about fulfilling needs withminimal material use and emissions. Thesustainability of PSSs depends on whether aPSS as such is less material intensive, andwhether actors in the chain feel incentives to lower material intensity even more. Thesecond question to be answered is whichfactors determine whether a PSS business modelper se generates less material flows and emissionsthan the competing product oriented models, and thus provides incentives for sustainablebehaviour?

This paper will first discuss what a PSS is, andpresent eight archetypical PSS business models.It will then present a theory about value cre-ation and a conceptual model about the rela-tion between PSSs and sustainability. Usingthese two models, the market potential andsustainability potential of the archetypical PSSmodels will be evaluated. The interest lyingwithin the overall analysis is then to identifysituations where both the business drivers andthe sustainability incentives for PSS are mutu-ally reinforcing.

2 For instance: an electronics firm does not want to devote energyand management time to running their company restaurant, andhence outsource it to a caterer. Many consumers with busy jobs nolonger bother buying potatoes, groceries and meat, and cookingtheir own meal, but rely on convenience food that can be servedin a few minutes.3 This is not so much an issue for someone who just sells a product:costs in the use phase are hidden and in a commodity market com-petition is on (lowest) sales price – if needs be by a poor designthat results in high use phase costs.4 The network is co-ordinated by TNO, econcept, VITO, CfSD,INETI and O2, and has some 30, mostly industrial, organizationsas direct participants. Furthermore, some 100+ other professionalsare involved. For more information, see www.suspronet.org.

5 It is maybe not totally coincidental that such ‘idealistic’ PSSs aremainly proposed by university researchers with hardly any firsthand experience in business development in companies. Kazazian(2002) tried to deal with this by asking for comments on his sus-tainable business ideas from real-world companies. Unfortunately,in most cases the latter were quick to point out the reasons whyKazazian’s ideas were not attractive businesses for them.

A. TUKKER


248

PRODUCT–SERVICE SYSTEMS:EIGHT ARCHETYPICAL MODELS

Various classifications of PSS have been pro-posed (see e.g. Behrend et al., 2003; Brezet et al.,2001; Zaring et al., 2001). Most classificationsmake a distinction between three main cate-gories of PSS. These are shown in Figure 1.

• The first main category is product-orientedservices. Here, the business model is stillmainly geared towards sales of products,but some extra services are added.

• The second main category is use-oriented ser-vices. Here, the traditional product still playsa central role, but the business model is not geared towards selling products. Theproduct stays in ownership with theprovider, and is made available in a differ-ent form, and sometimes shared by anumber of users.

• The last main category is result-oriented services. Here, the client and provider inprinciple agree on a result, and there is nopre-determined product involved.

However, each category itself includes PSSswith quite different economic and environ-mental characteristics. Elaborating on a typol-

ogy developed in a Dutch PSS manual (Tukkerand van Halen, 2003), it is possible to identifythe following more specific PSS types.

Product-oriented services

• Product-related service. In this case, theprovider not only sells a product, but alsooffers services that are needed during theuse phase of the product. This can imply, forexample, a maintenance contract, a financ-ing scheme or the supply of consumables,but also a take-back agreement when theproduct reaches its end of life.

• Advice and consultancy. Here, in relation tothe product sold, the provider gives adviceon its most efficient use. This can include, forexample, advice on the organizational struc-ture of the team using the product, or opti-mizing the logistics in a factory where theproduct is used as a production unit.

Use-oriented services

• Product lease. Here, the product does notshift in ownership. The provider has owner-ship, and is also often responsible for main-tenance, repair and control. The lessee pays

Value mainly in product content

Valuemainly inservicecontent

Product-service system

Product content (tangible)

Service content (intangible)

Pure Product

A: Productoriented

B: Useoriented

C: Resultoriented

Pureservice

1. Product related

2. Advice and consultancy

3. Product lease

4. Product renting/ sharing

5. Product pooling

6. Activity ma-nagement

7. Pay per service unit

8. Functional result

Figure 1. Main and subcategories of PSS

a regular fee for the use of the product; inthis case normally he/she has unlimited andindividual access to the leased product.

• Product renting or sharing. Here also, theproduct in general is owned by a provider,who is also responsible for maintenance,repair and control. The user pays for the useof the product. The main difference toproduct leasing is, however, that the userdoes not have unlimited and individualaccess; others can use the product at othertimes. The same product is sequentially usedby different users.

• Product pooling. This greatly resemblesproduct renting or sharing. However, herethere is a simultaneous use of the product.

Result-oriented services

• Activity management/outsourcing. Here a partof an activity of a company is outsourced toa third party. Since most of the outsourcingcontracts include performance indicators tocontrol the quality of the outsourced service,they are grouped in this paper under result-oriented services. However, in many casesthe way in which the activity is performeddoes not shift dramatically. This is reflectedby the typical examples for this type, whichinclude, for example, the outsourcing ofcatering and office cleaning that is now acommonplace in most companies.

• Pay per service unit. This category contains anumber of other classical PSS examples. ThePSS still has a fairly common product as abasis, but the user no longer buys theproduct, only the output of the productaccording to the level of use. Well knownexamples in this category include the pay-per-print formulas now adopted by mostcopier producers. Following this formula,the copier producer takes over all activitiesthat are needed to keep a copying functionin an office available (i.e. paper and tonersupply, maintenance, repair and replace-ment of the copier when appropriate).



249

• Functional result. Here, the provider agreeswith the client the delivery of a result. Thiscategory is used in this article, in contrast to activity management/outsourcing, for afunctional result in rather abstract terms,which is not directly related to a specifictechnological system. The provider is, inprinciple, completely free as to how todeliver the result. Typical examples of thisform of PSS are companies who offer todeliver a specified ‘pleasant climate’ inoffices rather than gas or cooling equipment,or companies who promise farmers amaximum harvest loss rather than sellingpesticides.

Going from the first to the last of these eighttypes of PSS, the reliance on the product as thecore component of the PSS decreases, and theneed of a client is formulated in more abstractterms. Every time the provider has a little morefreedom in fulfilling the true final need of theclient. However, abstract demands are oftendifficult to translate into concrete (quality per-formance) indicators, which makes it difficultfor the providers to determine what they haveto supply, and difficult for the clients to knowwhether they have got what they asked for.

THE BUSINESS CASE FOR PSSs:FACTORS INFLUENCING ADDEDVALUE AND INNOVATIONPOTENTIAL

This section deals with the first key question ofthis paper: which factors determine whether a PSSbusiness model is the best way to create valueadded?6 To address this question, the factors

6 In this I assume a fairly rational approach of firms to business. Ofcourse sometimes bounded rationalities and existing heuristicswill prevent a firm from seeing easily achievable win–wins (cf.Simon, 1957; Lindblom, 1959), but usually the pros and cons ofembarking on a specific business model are well analysed, andpreferences for incremental innovation trajectories all too often canbe explained as rational and cost-effective building upon previousinvestments in fixed assets, equipment, human resources etc. (cf.Dosi, 1982).

that determine whether added value is at stakeor not are analysed, and the situation is thenreviewed for each PSS business model.

Value elements in PSSThe ability to create and capture sustainedadded value (often referred to as shareholdervalue) is often seen as the key measure ofsuccess of business. In this paper I use theconcept of economic value added (or EVA) as pro-posed by Stewart (1991). He defines EVA as

A fundamental measure of corporate per-formance, it is computed by taking thespread between the return on capital andthe cost of capital, and multiplying by thecapital outstanding at the beginning of theyear [. . .]. It is the residual income thatremains after operating profits cover a fulland fair return on capital (i.e. the cost ofcapital). In theory, a company’s marketvalue added at a point in time is equal tothe discounted present value of all EVA itcan be expected to generate in the future [. . .].

On the basis of the definition above, a numberof key elements can be distinguished:

• market value of the PSS (tangible and intangible);

• production costs of the PSS (including riskpremium aspects);

• investment needs/capital needs for PSSproduction and

• the ability to capture the value present in thevalue chain, now and in the future.

The difference between the first and secondelements, in combination with the fourth,determines the operating profits of the PSSprovider (network) – now, and in the future.The third element indicates whether a need forhigh or low capital is a factor in determiningsuch profits. This is highly relevant, since if thesame profit can be made with lower capitalneeds this leads to a higher EVA.

A. TUKKER


250

The four elements described above need tobe discussed in more detail. The aim is to comeup with a set of criteria that helps to evaluatewhether it is better to put a PSS on the market,compared with a product.7

Market value of the PSS (tangible and intangible)

This element includes the following aspects.

• Tangible or objective value for the consumer(e.g. resources, time input and cost of capitalsaved).

• Intangible or subjective value for the consumer (e.g. additional, ‘priceless’ experiences).

Tangible or objective value is a fairly straight-forward concept. A customer who has thechoice between buying a product, or using aPSS, can start to make a rational calculationabout what the product actually costs, includ-ing all kinds of ‘hidden’ cost, and that is inprinciple the maximum price he/she wouldlike to pay for a competing PSS. For instance,in the concept of chemical manufacturing ser-vices the PSS provider not only sells chemicals,but takes over specific handling and manage-ment tasks for the client. This saves the clienthuman resources, management time andmaybe even space, which often costs five to tentimes more than the chemicals themselves(Votta, 2003). Intangible or subjective value isa little less straightforward as a concept, but iscurrently the key to success or failure of manyproducts and services in the consumer market.In an affluent (Western) society consumers cangenerally take basic Maslowian needs such asfood, shelter and safety for granted, and willbe more geared toward the realization ofhigher needs such as affiliation, love, esteemand self-realization. Authors such as Pine and

7 The comparison is done per functional amount product and PSS.This is probably an oversimplification, since it is likely that successon the market will differ and hence market volume too. In princi-ple, one could try to expand the line of thought developed hereinto a quantitative evaluation model, but that is an endeavour Ileave for another time.



251

Gilmore (1999) and LaSalle and Britton (2003)argue that society is shifting to an experienceeconomy (see also Grinyer, 2003).8 The trick thenbecomes to satisfy needs on these higher levelsin conjunction with the offer of a material arte-fact: ‘turning ordinary products into extraordi-nary experiences’.9 By creating such intangibleadded value, the provider makes the clientwilling to pay more than would be justified onthe basis of ‘rational’ calculation. This allowsthe provider to charge more, but the user isforced to diminish other expenditures to paythe extra.

Production costs of the PSS (including riskpremium aspects)


• Traditional ‘tangible’ production costs (e.g.resources, time input and cost of capitalused).

• Risk premium/uncertainty related to thesolution.

Traditional tangible production costs oftenmirror the tangible or objective value for theconsumer discussed above. Often, servicingimplies that activities formerly performed bythe user are now performed by the provider.This makes sense if the tangible added valueof the PSS for the user is higher than the (extra)production costs for the provider, or if a costdeficit is more than compensated for by theintangible added value that the user attributesto the PSS.10 Risk premium or uncertainty is

relevant for (specifically) result-oriented PSSs.By promising a result, the provider often facesdifficulties in predicting and controlling risks,uncertainties and responsibilities that other-wise were the problem of the user. Forinstance, if a company sells pesticides andgives advice on how to use them, the farmerbears the full risk of success of crop growth. Ifthe provider promises a result such as an x%reduction in loss of crop due to pests, a heftypenalty might have to be paid if the providerfails.11 If such liabilities cannot be controlled,the provider should either refrain from such aPSS or include a (probably prohibitive) high-risk premium in its cost price.

Inherent capital needs plus investment needs forPSS production


• The inherent capital base needed to producethe solution.

• Any additional ‘transition’ investmentsneeded to create the system that producesthe solution.

The first aspect in principle concerns a straight-forward comparative calculation of the capitalneeds of the technical and organizational struc-ture of the product and PSS providing systems.There can be important differences betweenPSS types. For instance, a company leasing carsneeds one car for each client. A company oper-ating a car pooling system can have consider-ably fewer cars for the same client pool, andhence needs to invest considerably less. Addi-tional transition investments relate to theassumption that in most cases companies arecurrently quite product oriented. They have toinvest in infrastructures and relationships to beable to provide a PSS. Such transition costs canbe of intangible nature, such as the problems

8 See, for example, the interesting overview of Vlek et al. (1999),who discuss need hierarchies in the context of sustainable con-sumption. Not totally coincidentally, LaSalle and Britton argue thathuman beings do not desire goods in themselves, but the benefitsgoods provide at the higher levels at which human beings operate,such as physical, emotional, intellectual and spiritual.9 As LaSalle and Britton subtitled their book. The main title, Price-less, was taken from a Mastercard commercial, which goes roughlyas this: Matches: 10 cts. Candles: $8. Relaxation in the atmospherecreated: priceless.10 A change from products to PSS might imply a shift between costcategories (e.g. lower costs for capital goods since they can be moreintensively used, but extra costs due to more complicated billingsystems and more service personnel etc.).

11 This is an example where one can define still meaningful andquite tangible indicators measuring success. If such indicatorscannot be found, putting a result-oriented PSS on the marketbecomes virtually impossible, or if one does, a source of quarrelbetween provider and client.

A. TUKKER


252

involved in breaking up a relationship with apreferred supplier. In principle such transitioncosts have to be treated as investments thattherefore tie in capital.

The ability to capture value present in the valuechain, now and in the future


• Strategic position in the value network:ability to capture value.

• Sustained low barriers for access to theservice and a contribution to client loyalty.

• Contribution to a comparatively high speedof innovation.

The first aspect has been stressed by Porter(1985) amongst others. The creation of (tangi-ble and intangible) value alone is not sufficient.The PSS provider (network) should be power-ful enough to capture this value as well. To dothis, the provider network has to create aquasi-monopoly by covering the essentialparts of the delivery or production system, i.e.the parts that cannot easily be copied or per-formed by other parties (e.g. unique relation-ships with clients, unique technologies, etc.).12

The second aspect is often associated with use-oriented and result-oriented services. First, thepayment reference is not for the product(which often has to be paid upfront, thusposing quite a barrier for the user), but apayment per unit time or unit use. Second,such a PSS offers a better means of ensuringclient loyalty than the mere purchasing of

products.13 The third aspect is related to thefact that PSSs might contribute to better andfaster innovation. This is in part related tomaintaining low barriers to accessing theservice, which also implies access to, andinsight into, the needs of clients.14 However,other aspects play a role. As indicated byRoome and Clarke (2002) amongst others,openness, predisposition to mutual learningand access to networks of other players todiscuss strategy are key features of innovativefirms. Since PSSs are so user oriented, and ingeneral put on the market through intense co-operations within networks of firms, it couldbe argued that this helps firms (or the networkas a whole) to be innovative.

Analysis of key economic elements per type of PSSTable 1 scores the eight generic types of PSSagainst the four key economic criteria asdescribed above in the four elements and theiraspects. Of course, for individual PSSs a tai-lored and quantitative evaluation is necessaryto find out whether it is profitable business.

Table 1 shows the following.

• Product-related services (1) and advice and con-sultancy (2) usually provide some tangiblevalue for the user by a more efficient use of materials and human resources. This isreflected in some additional material andhuman resource costs for the provider. Aproduct-oriented company embarking onthese types of PSS usually has to make some

12 Christensen et al. (2001) give the dramatic example of IBM. In theearly 1980s, this firm covered some 70% of the whole value chainof (main frame) computers and attracted 90% of its added value.Squeezing out less powerful players in the chain seemed a prof-itable strategy, and IBM started to outsource part of their non-essential activities to enhance return on investment even more.However, one decade later they discovered their mistake. The keystrategic bottlenecks in PC production and use were now the oper-ating system (monopolized by Microsoft), and the processor(monopolized by Intel, and, at a distance, AMD). Anyone couldbuild a PC around these essential elements, with no effective wayof being really distinctive in design. As a consequence, the powerin the value chain shifted dramatically, the processor suppliers andMicrosoft capturing most profits, and for a while IBM was in sometrouble. See also Raffii and Kampas (2002) on this topic.

13 This can take a negative form, e.g. if by signing a contract theclient can no longer shop around with other suppliers of consum-ables and service; but also a positive form: supplying a PSS helpsto build a unique client–provider relationship and helps to give theprovider unique insight into specific client needs, enabling tailor-ing of its services more and more over time to this specific client,which in turn creates a situation where competitors are not able tocome up with a competing offer unless they make a considerableinvestment in building a similar relationship.14 Think of the well known story that, since copier producers workwith a business model based on pay per print, their maintenancestaff regularly visits clients and can report back how the copier isused, which can be directly translated into the design or adapta-tion of copiers.



253

investments in capital and organizationaltransitions. There might be a benefit in termsof lower client barriers (if financial servicesare included), a higher client loyalty and,due to better client contacts, some increasein the speed of innovation.

• Product lease (3) has some tangible value forthe user, since various costs and activitiesare shifted to the provider. The providermight have to make provisions for morecareless client behaviour. Since the providerremains as owner of the product, the needfor capital is high. Barriers to attracting newclients are low due to low initial investmentby the client. User loyalty might improve (asthe product plus maintenance, etc. is pro-vided), but the user can still easily switch to

other providers. Since leasing companiesuse products provided by others, no influ-ence on innovation is assumed.

• Product renting and sharing (4) in generaldemands a tangible sacrifice by the user.He/she now has to put time and effort intogetting access to the material artefact. Com-pensation can come from the fact thathe/she no longer needs to bear the capitalcosts of the product. It is likely that this PSStype scores low in terms of intangible value.Rental equipment in many cases does notcontribute to (self-)esteem, or ‘priceless’experiences, though there are exceptions(‘rent this BMW and be a king for a day’).The organizational system at the provideruses more input of human resources. Since

Tang

ible

val

ue

Inta

ngib

le v

alue

Tang

ible

cos

ts

Ris

k pr

emiu

m

Inhe

rent

cap

ital

Tran

siti

on c

osts

% V

alue

cap

ture

d

Clie

nt lo

yalt

y

Hig

h sp

eed

of

inno

vati

on

Mar

ket

valu

efo

r th

e us

er

Cos

ts f

orpr

ovid

er

Cap

ital

nee

ds

Abi

lity

to s

usta

inva

lue

in f

utur

e

Table 1. A discussion of the value characteristics of different types of PSS

PSS type

1. Product-related service 0/+ 0/+ -/0 0 -/0 -/0 0 + 0/+2. Advice and consultancy 0/+ 0/+ 0 0 0 -/0 0 + 0/+3. Product lease 0/+ ? -/0 -/0 - - + -/+ 04. Product renting or sharing -/+ –– + 0 -/+ - ? ? ?5. Product pooling -/+ –– + 0 -/+ - ? ? ?6. Activity management +? +? 0 -/0 0 - ? + +7. Pay per unit use + ? 0 ? 0 - + + +8. Functional result 0 ? ++/? –– ++/? –– ? ? +

Key++: Much better than reference (product)+: Better than reference0: Indifferent–: Worse than reference––: Much worse than reference?: No judgement possibleGrey: Most problematic areas

the provider keeps on owning the product,capital need is high. However, due to theshared use, overall capital need in thesystem is considerably lower. Due to lowinitial costs the access barrier for new clientsis low.

• Product pooling (5). The analysis is virtuallythe same as for renting and sharing.

• Activity management (6). Activity manage-ment shifts personnel and material costsfrom the user to the provider, who has tomake gains by organizing the outsourcedtasks more efficiently by specialized knowl-edge. It is important that good performancecriteria can be defined, since otherwise dis-cussion about the delivered result can arisebetween user and provider (risk premiumissue). Since activity management is usuallyarranged via longer term contracts, reason-able client loyalty is ensured. The special-ization might lead to a high speed ofinnovation.

• Pay per unit use (7). There is a clear tangiblevalue for the user since various activities(maintenance, etc.) are outsourced to theprovider. The providers’ position in thevalue chain becomes better, in relation todirect access to clients and (enforced) clientloyalty. The provider has to be able topredict the behaviour of the user, since otherwise no clear cost calculation can bemade and a risk premium has to beincluded. Since the product stays in owner-ship of the provider additional capital isneeded. There are low barriers for newclients, and good client contacts in principlelead to better innovation potential.

• Functional result (8). Since the same functionis offered, in principle the user could give it the same tangible value. Intangible valueis another matter though, and cannot bejudged without defining the specific system.In principle, the provider could try toprovide a solution with much lower input ofhuman resources and materials. However,since the provider promises a result on ahigh level of abstraction, agreement on per-

A. TUKKER


254

formance indicators, and the level of controlin achieving this performance, can be animportant (if not prohibitive) problem(which translates to a low score on the riskpremium issue). Capital costs could be low,but transition costs high. This model leavesthe highest degree of freedom with regard toinnovation.

The overall picture is that product-orientedservices are the least radical ones and probablyeasily applicable by traditional product-oriented firms. Use-oriented services are nowalso common business models, whereasproduct renting, pooling and sharing seems tohave a relatively high chance of creating tangi-ble and intangible client sacrifices. Withinresult-oriented services, activity managementand pay per use are becoming more common.The key problem with these PSSs is the diffi-culty of agreeing with the user a set of goodperformance criteria, and the prediction of, orinfluence on, the behaviour of the user withinreasonable margins. This risk element is par-ticularly relevant for the functional result typeof PSS, since the provider takes over all the lia-bilities that in a product-based system werewith the user.

The sustainability case: factors influencingsustainability potentialIn this article the sustainability of PSS typeshas been analysed with the help of ‘sustainabledesign rules’ (see Manzini, 2002; Zaring et al.,2001; McDonough and Braungart 200115). Thefollowing reasons for differences in environ-mental performance of products and PSS havethus been identified (including a tentative indi-cation of the magnitude of environmentalimpact reductions).

15 Typically, PSSs can result in lower impacts than products by theirinherent design, or by stimulating user and provider behaviourgeared towards low material and energy use. Quantitative or semi-quantitative approaches such as LCAs can only be applied to con-crete PSS cases and are less suitable in this analysis. Please notethat in this article the issue of sustainability focuses on environ-mental sustainability only.

In the next sections, the effects of each type ofPSS are discussed, and summarized in Table 2.

Product-related service (1)

The majority of product-oriented PSSs do notimply any change in the technological systemor how the user operates it (relating to thethird, fourth, fifth and sixth mechanisms iden-tified above). There is no strong incentive interms of internalizing true life cycle costs in thedesign process by the provider. There might besome incremental efficiency improvementsdue to better maintenance, or take-back provi-sions, although even these might be absent.The user may organize good levels of mainte-nance by hiring third parties, and take-backschemes might already be legally demanded.Financing schemes have no effect whatsoever.The overall picture is that product-related ser-vices can lead to impact reductions, but theyare likely to be incremental at best.

Advice and consultancy (2)

The effects here are similar to those forproduct-related services. Again, the mainvalue lies in the fact that the PSS providermight suggest all kinds of optimization forusing the product, which in the end can leadto incremental reductions in environmentalimpacts.

Product lease (3)

In the case of product lease, it is not clearwhether there will be impact reductions or not.In principle, the provider now also takesresponsibility for maintenance, repair andcontrol, and this could lead to incremental effi-ciency improvements: the product has a some-what longer life span and might use energyand consumables more efficiently by bettermaintenance, repair and control. In principle,the provider may perceive an incentive toprolong the product life and may design theproduct accordingly. In most cases, however,lease companies buy the products they lease



255

Mechanisms leading to incremental/averageimpact reductions (10–20%)

• Incremental efficiency improvements (e.g.by better maintenance due to a maintenancecontract in a product-related service). Thiscan lead to a more intensive use or a pro-longed life of capital goods, or less use ofenergy and consumables in the use phase –though in a less radical manner as indicatedin the second, third and fourth reasonsdescribed below.

Mechanisms leading to average to high impactreductions (up to 50% or factor 2)

• Designing products taking true life-cyclecosts into account by the provider, whotakes responsibility for these life cycle costs(e.g. in the case of pay per service unit). Thisis a strong incentive for optimizing the useof energy and consumables, and recycling ofproduct parts and materials where feasible.

• Considerably more intensive use or pro-longed life of capital goods used in thesystem (e.g. in a product renting or sharingsituation) than in a traditional productsystem. Intensive capital good use can alsostimulate a quicker replacement by newer,more efficient models.

• Considerably less use of energy and otherauxiliary materials in the use phase, e.g. in aproduct pooling situation. The same usephase in energy and auxiliary material useis shared by various users.

• Use of a considerably more efficient tech-nology made possible by a higher economyof scale (e.g. washing machines in a laundryusing gas heated water rather than the elec-tricity heated water used at home).

Mechanisms potentially leading to very highimpact reductions (up to 90% or factor 10)

• Application of a radically different techno-logical system with radically lower impacts(e.g. a functional result).

A. TUKKER


256

for third parties, and are not responsible forproduct design. So the incentive felt by theleasing company is not passed on to a designerteam directly. Furthermore, the lease in generaldoes not cover many costs in the use phase(e.g. fuel consumption in cars), so neither thelessee nor the product provider will perceivemuch incentive to do something about energyand consumable use in the use phase. The factthat the user no longer owns the product couldeven lead to negative effects, such as a carelessuse shortening its useful life span.

Product renting and sharing (4)

Product renting and sharing implies that sameproduct is now more intensively used. This canhave high impact reductions, particularly if thelife-cycle impacts are mainly related to themanufacture of the product (compare Wimmerand Züst, 2001). This PSS can have an addi-tional bonus. In general, the user will nowhave to pay the integrated costs for each time

he/she uses the product, unlike the case for theformer PSS types. Also, access to the productis a little more complicated. This implies thatin this system the use of the product in generalwill be somewhat discouraged. This might haveadditional positive environmental effects, if itleads to a less-use situation, or to more fre-quent use of more environmental friendlyalternatives (e.g. public transport as a comple-ment to car renting or sharing).

Product pooling (5)

The analysis for product pooling is similar tothat for renting and sharing, with one majordifference. Product pooling implies that thesame product is used at the same time by moreusers (e.g. car pooling). This can have evenmore impact reductions than in the case ofsharing and renting, particularly if the life-cycle impacts are related to the use of theproduct.

Table 2. Relevance of impact reduction mechanisms per PSS type

Mechanism Relevance per PSS type

With incremental impact reductionIncremental efficiency improvements + + ? ? ? + + +With average to high impact reductionDesign based on true life cycle costs ?. n.r. ?. ?. ?. +/? + +More intensive use/prolonged life capital goods n.r. n.r. n.r.. + + +/?. +. +Much less use of energy etc. in use phase n.r. n.r. n.r. + + +/? ? +Much more efficient technology by economy of scale n.r. n.r. n.r. n.r. n.r. ? n.r. ?With very high impact reductionsApplication of a radically different, low impact technology n.r. n.r. n.r. n.r. n.r. n.r. n.r. +

Remark: product renting/sharing and pooling have a bonus in that they discourage use of products.

1.Pr

oduc

t re

late

d

2.A

dvi

ce a

nd c

onsu

ltan

cy

3.Pr

oduc

t le

ase

4.Pr

oduc

t re

ntin

g/sh

arin

g

5.Pr

oduc

t po

olin

g

6.A

ctiv

ity

man

agem

ent

7.Pa

y pe

r us

e

8.Fu

ncti

onal

res

ult



257

Activity management/outsourcing (6)

Activity management or outsourcing usuallydoes not imply a radical change in appliedtechnology, organization etc. However, com-panies providing this PSS (financially) have tobe more efficient than the company who out-sourced the activity to stay in business. Thiscan be realized by a more efficient use ofcapital goods and materials, although it seemsunlikely that the gains are more than a fewdozen per cent. On top of this, in many casesthe efficiency gains are realized on personnelcosts rather than material costs, which is lessrelevant for impact reduction.

Pay per unit use (7)

Two aspects concerning the pay per unit of usePSS are of relevance. First, the provider isresponsible for all life cycle costs, which pro-vides a powerful incentive to design a productthat in terms of costs is optimized over the lifecycle, of which elements can be re-used afterthe products’ useful life. Second, in specificcases (e.g. pay per wash) the user will make amore conscious use of the service, though inother cases (e.g. copiers at work) this issue

plays no role. A very important issue is that the provider feels an incentive to continuallyimprove the product with life-cycle perfor-mance in mind.

Functional result (8)

Offering a functional result in principle has thehighest potential for impact reduction. At ahigh level of abstraction, a result is promisedand the provider can decide the necessaryapproach to deliver the result. This providerwill therefore try to do so in the most cost-effective way, which bears the promise of asearch for radical innovations.

Table 3 below tentatively translates the find-ings of Table 2 to give an impression of theenvironmental sustainability potential of thedifferent types of PSS.

The conclusions from this translation are the following.

• Product-related service, advice and consul-tancy, and product lease types of PSS haveprobably marginal environmental benefits,since at best mainly incremental changesuch as better maintenance etc. can beexpected. Product lease PSS can even

Table 3. Tentative (environmental) sustainability characteristics of different PSS types

PSS type Impacts compared to reference situation (product)

Worse Equal Incremental Considerable Radicalreduction reduction reduction

(<20%) (<50%) (<90%)

1. Product-related service ¨- - - - - - - - - - - -Æ2. Advice and consultancy ¨- - - - - - - - - - - -Æ3. Product lease ¨- - - - - - - - - - - - - - - - - - - - - -Æ4. Product renting and sharing ¨- - - - - - - - - - - - - - - - - - - - - - - - - - - - -Æ5. Product pooling ¨- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -Æ6. Activity management ¨- - - - - - - - - - - - - - - - - - - - - - -Æ7. Pay per unit use ¨- - - - - - - - - - - - - - - - - - - - - - - - - - - - -Æ8. Functional result ¨- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -Æ

Notes:• Renting, sharing: radically better if impact related to product production.• Pooling: additional reductions compared with sharing/renting if impacts related to the use phase.• Renting, sharing, pooling: even higher if the system leads to no-use behaviour.

produce negative environmental effects ifthe lessee has incentives to use the productin a less careful way than when he/she ownsit.

• Product renting, sharing or pooling PSSs canhave major environmental benefits if theburden is related to the production of theartefact, since the same product is sharedand used more intensively. However, if, inthe case of product renting or sharing, theuse phase dominates and does not lead notto a low use behaviour there is little positiveoutcome. In such cases, pooling leads tolower impacts since more people make useof the product at the same time.

• An activity management/outsourcing PSSwill lead to lower environmental impacts if(monetary) efficiency gains are particularlyrelated to materials and artefacts, and not totime input of humans.

• The pay per unit use PSS overcomes the splitincentive between production costs of aproduct and costs made in the use phase. Itis likely that at least incremental gains willbe realized, but since the technologicalsystem in principle does not change radi-cally, no radical improvements can beexpected.

• Functional result PSSs have, in theory, thehighest potential since the provider offers aresult closer to a final client need and hencehas more degrees of freedom to design alow-impact system.

Overall, it can be cautiously concluded thatmost PSSs will probably lead to some environ-mental improvements, or at least no worseenvironmental performance. The exception isformed by PSSs that make users less responsi-ble for careful use of the product (leasing).Improvements tend to be incremental toaverage at best, though. Such improvementsare mainly related to economic efficiency gains,and might therefore be less relevant forhuman-resource intensive systems. Radicalimprovements can only be expected in the caseof promising functional results.

A. TUKKER


258

CONCLUSIONS

The overall conclusions of this paper are notstraightforward. True, in comparison toproduct systems most PSSs analysed havesome potential for environmental gains, and ingeneral there is not too much risk that addi-tional impacts will be created.16 However, ithas also become clear that the PSS in general isnot a panacea for reaching radical factor 4 or10 environmental improvements. Of course, intheoretical terms it is easy enough to define afunctional result at a high level of abstraction(e.g. a person that needs transport from A to B),and then to conclude that systems can bedesigned that can deliver this function a factorof 4–10 times more efficiently than the cur-rently dominant product system, i.e. transportby own car. However, it does not work theother way around. Most PSS types cannot be expected to result in such radical gains.Further, the conclusions reached from ananalysis of the business and sustainabilitycases for PSS show that there are a number ofinteresting contradictions concerning whatwould be desirable from a sustainability pointof view, and what makes good sense for business. The simple thinking that PSS development will automatically result in anenvironmental–economic win–win situationalso seems to be a myth. Some of the contra-dictions include the following.

• The least problematic PSSs for companies to introduce (in many cases still product oriented) are product-related services andadvice and consultancy. These lead at best tosome incremental environmental improve-ments. The same applies, though maybe to a lesser extent, to activity management.Radical changes cannot be expected sincethe technological system itself basically doesnot change, and is merely operated moreeffectively.

16 Abstracting from rebound effects that might be created if PSS ismore cost-efficient than the product, and the saved income is spenton other activities.



259

• Some PSSs, such as product lease, can leadto less responsible user behaviour and hencean increase of environmental impacts.

• Product renting, sharing and pooling inprinciple can lead to high environmentalgains: capital goods are used more inten-sively and, in the case of pooling, consum-ables in the use phase are now beneficial formore persons at the same time. However,this category in particular probably has aconsiderably lower market value than thecompeting product, due to both tangible andintangible user sacrifices.17

• The most promising PSS in environmentalterms is the function-oriented PSS.However, this PSS particularly needs atten-tion concerning operationalization. Therewill be plenty of cases where the ‘functionalresult’ cannot be operationalized in suffi-ciently concrete terms, where liabilitiesrelated to the promised result are too high,or where the provider simply has insuffi-ciently control if the result is to be reached.These issues can be prohibitive in putting a function-oriented PSS on the market (orotherwise only by demanding a high riskpremium).

These findings, incidentally, also indicate someimportant issues for a research agenda thataims at promoting sustainable PSSs. It is all tooclear that product renting, sharing, poolingand functional PSSs are probably the mostpromising types from a sustainability point ofview. Approaches should be developed thatovercome the main barriers to implementationfor such PSSs, i.e.

• design of product renting, sharing andpooling systems that have a high intangiblevalue for the user, while sacrifices withregard to tangible value are minimized;

• development of concrete performance indi-cators for functional PSSs;

• development of approaches which canreduce the liability risks and enhance controlover ‘production’ uncertainties related tofunctional PSSs.

REFERENCES

Behrend S, Jasch C, Kortmap J, Hrauda G, Firzner R, VelteD. 2003. Eco-Service Development. Reinventing Supply andDemand in the European Union. Greenleaf: Sheffield.

Brezet JC et al. 2001. The Design of Eco-Efficient Services;Method, Tools and Review of the Case Study Based ‘Design-ing Eco-Efficient Services’ Project. Ministry ofVROM–Delft University of Technology.

Christensen FCM, Raynor M, Verlinden M. 2001. Skatewhere the money will be. Harvard Business ReviewNovember: 72.

Dosi G. 1982. Technological paradigms and technologicaltrajectories. Research Policy 11: 147–162.

Grinyer C. 2003. Presentation on the business model ofOrange UK during the Launch Conference of Sus-Pronet, Amsterdam. Available via Orange UK, London,or www.suspronet.org

Kathalys. 2002. Vision on Sustainable Product Innovation.p/o TNO Industrial Research, Eindhoven (see alsowww.kathalys.org).

Kazazian T for WWF France. 2002. Il y Aura l’Age desChoses Legeres. Design and Developpement Durable [TheAge of Light Things is Arriving. Design and SustainableDevelopment]. Victoire: Paris.

LaSalle D, Britton TA. 2003. Priceless. Turning OrdinaryProducts into Extraordinary Experiences. Harvard Busi-ness School Press: Boston, MA.

Lindblom CE. 1959. The science of muddling through.Public Administration Review 19: 79–99.

Manzini E. 2002. Sustainable Solutions. New Business Ideasand New Ideas on Business, working paper, INDCAO,CIR.IS, Politecnico di Milano.

McDonough W, Braungart M. 2001. The next industrialrevolution. In Sustainable Solutions, Charter M, TischnerU (eds). Greenleaf: Sheffield; 139–150.

Montalvo Corral C. 2002. Environmental Policy and Tech-nological Innovation. Why do Firms Reject or Adopt NewTechnologies? Elgar: Cheltenham.

Pine JB II, Gilmore JH. 1999. The Experience Economy.Harvard Business School Press: Boston, MA.

Porter M. 1985. Competitive Advantage. Creating and Sus-taining Superior Performance. Free Press: New York.

Rafii F, Kampas PJ. 2002. How to identify your enemies –before they destroy you. Harvard Business ReviewNovember: 115.

17 For instance, recently it became clear that one of the main spon-sored car sharing programmes in the Netherlands only had a fewthousand subscribers and has no major growth in view. It is all tooclear that for the vast majority of the Dutch population suchschemes are not a realistic and attractive alternative to owning orleasing an ‘own’ car. The subsidy of the car sharing organizationis now in discussion.

Votta T. 2003. Chemical Manufacturing Services. Presenta-tion during the Launch Conference of SusPronet, Amsterdam. Available via the Tellus Insitute, Boston,US, or www.suspronet.org

Wimmer W, Züst R. 2001. Ecodesign-Pilot. Orell Fuessli:Zurich.

Zaring O (ed.). 2001. Creating Eco-Efficient Producer Ser-vices, report of an EU project. Gothenburg ResearchInstitute, Gothenburg.

BIOGRAPHY

Arnold Tukker (Project Manager SusProNet),Programme Manager Sustainable Innovation,TNO Institute of Strategy, Technology andPolicy, P.O. Box 6030, 2600 JA Delft, TheNetherlands.Tel.: + 31 15 269 5450Fax: + 31 15 269 54 60E-mail: [email protected]

A. TUKKER


260

Roome N, Clarke S. 2002. Sustainable business: learning-action networks as organisational assets. In Partnershipand Leadership, De Bruijn Th, Tukker A (eds). Kluwer:Dordrecht.

Simon HA. 1957. Theories of decision making in eco-nomics and behavioural science. American EconomicReview 49: 253–283.

Stewart GB. 1991. The Quest for Value. The EVA Manage-ment Guide. Harper Business: US.

Tischner U, Verkuijl M, with Tukker A. 2002. First DraftPSS Review. SusProNet Report, draft 15 December. Avail-able from Econcept, Cologne, Germany; TNO-STB,Delft, the Netherlands, or www.suspronet.org

Tukker A, van Halen C (eds). 2003. Innovation ScanProduct Service Combinations, manual. English versionavailable in October 2003 from TNO-STB, Delft, orPricewaterhouseCoopers, Utrecht, the Netherlands.

UNEP. 2002. Product Service Systems and Sustainability.Opportunities for Sustainable Solutions. UNEP–DTIE:Paris.

Vlek CAJ, Rooijers AJ, Steg EM. 1999. Duurzamer Con-sumeren: Meer Kwaliteit van Leven met Minder Materiaal?[More Sustainable Consumption: More Quality of Life withLess Material?], research report for the Dutch Ministry ofEnvironment. COV, Groningen University, Groningen.

Documents

Eight types of product–service system: eight ways to sustainability? Experiences from SusProNet