incubation for growth

8/4/2019 Incubation for Growth

1/54

Research summary: September 2011

Incubation for GrowthA review of the impact of business incubationon new ventures with high growth potential

Nicola J. Dee, Finbarr Livesey, David Gill and Tim Minshall


2/54

NESTA is the UKs foremost independent expert on how innovation cansolve some of the countrys major economic and social challenges. Its work isenabled by an endowment, funded by the National Lottery, and it operatesat no cost to the government or taxpayer.

NESTA is a world leader in its eld and carries out its work through a blendof experimental programmes, analytical research and investment in early-stage companies. www.nesta.org.uk

Contents

Incubation for Growth

A review of the impact of business incubation on new ventures with high growth potential

Part 1: Introduction 3

Part 2: Summary of ndings 5

Part 3: Research background 8

Part 4: Evidence on the impact of business incubation 13

Part 5: Understanding the impact of business incubation on new ventures 17

Part 6: Matching incubator services to the changing needs of rms 25

Part 7: The incubator business model 32

Part 8: Concluding remarks 38

Appendix A: Performance indicators for business incubation 41

Appendix B: Review of quantitative academic contributions 45

Appendix C: Reference points for incubator business model 48

References 50

Acknowledgements 53


3/54

1.1 Overview of the research

The Government has put high-growth,innovative businesses at the heart of itseconomic agenda, and is focusing policy onhow to back the big businesses of tomorrow.The aim of this research was to provide: athorough and ocused literature review on

business incubation.The purpose of whichwas to identify models of incubation that havethe greatest impact on the mission of buildinghigh-growth, innovative rms. This involved:

Mapping the range of existing models

relevant to rms with high-growth potential.

Clarifying state-of-the-art (models,interventions, design principles) in businessincubation across the public and privatesectors.

Gathering, analysing and synthesisingexisting work (quantitative and qualitative)on the impact of incubation.

While the research was focused on the impactof incubation on tenants, we also examinedthe broader economic impact of businessincubation.

1.2 Approach

Business incubation includes a variety ofmechanisms and objectives as described. Wehave focused our work on business incubationdesigned to impact high-growth innovativebusinesses, and business incubators with

physical space. We have organised the reportbased on a review of the academic businessincubation literature, with additions from

industry reports. We have found signicantlimitations in the business incubation literature

which has led us to include, where possible,literature relating more generally to innovationand entrepreneurial activity.

There has been much confusion regarding thedenition and impact of incubation. Followingthe publication of two in-depth reviews ofresearch on incubators, (Hackett and Dilts2004b; Phan, Siegel et al. 2005), we havefocused on reviewing literature publishedduring the last ten years. A larger window forreview would not have been possible within the

budget constraints.

1.3 Structure of the report

Much of the literature on business incubationhas inconsistent, and at times, conictingmessages. We have sought to explain why thishas occurred, in addition to attempts to resolvesome of these difculties. The structure of thereport is as follows:

Summary of ndings: a high-level view ofndings.

Research background: including a reviewof existing recognised models of businessincubation, and more details on our approach.

A top line view: aggregate view on the impactof the business incubation industry.

Understanding the impact of business

incubation on new ventures: expected

outcomes and tools of incubation.

Part I: Introduction


4/54

4

Matching process: creating value by matchingincubator services to the needs of rms.

The incubator business model: while weprovide some guidance, further research isneeded.

Concluding remarks

Appendices: further information and ndings.


5/54

Business incubators have proliferated sincetheir emergence over 50 years ago. Over this

time business incubation has evolved to includea range of incubation practices. Nonethelessbusiness incubation can deliver critical valueto tenants. Contrasting early denitions ofincubation where survival of tenants wasemphasised, we dene incubation as ...a

shared ofce-space acility that seeks to

provide its incubatees with a strategic, value-

adding intervention system o monitoring and

business assistance.1 Our key ndings followthe structure of the report.

The proliferation of business incubation

over the last 50 years has resulted in

diversication of terminology used andtypes of incubation offered.To helpovercome this problem we compared businessincubation with two activities sometimesconfused with incubation equity nancingand professional services rms. For example,though not as intensive for venture capitalists,incubators implement an entry selectionprocess for tenants. Perhaps more importantlyincubators often have a very mixed revenuestream and incentives as a result, strongly

Part 2: Summary of ndings

1. Hackett, S. M. and Dilts,D.M. (2004b) A SystematicReview of Business IncubationResearch. Journal ofTechnology Transfer. 29:55-82.

Figure 1: Theoretical impact of an incubator on the irregular growth path of an individualtenant

Growth

Incubationp

eriod

Time

Old growth path New growth path


6/54

encourage peer-to-peer networking, addressmultiple needs of new ventures withoutprioritising just one, and offer continualexposure to the incubation environment andservices.

Absolute measures of incubation are

impractical, but performance indicators areuseful. Given the relatively small number ofstudies and the lack of comparability betweenthem, any conclusions should be treated asindicative at best.

The UK has approximately 300 businessincubators supporting around 12,000businesses (UKBI). Estimates of the directimpactof business incubation by industryassociations2 include between 25-40 supportedbusinesses and between 44-91 jobs perincubator. Many incubators (~60 per cent)

also have outreach programmes to supportbusinesses not resident in the incubator.

Indirectincubator effects, e.g. additionaljobs and wealth generation from providingproducts/services to incubator and incubatees,globally range between 0.48-1.5 times thedirect impacts of incubation.

Few studies capture the full impact of businessincubation, for example taking a measureof incubation impact over the incubation

period rather than longer term, and ignoringentrepreneurial learning and subsequentventuresome activity as a result of businessfailure. Job creation, while a popular metricused to evaluate incubation, is not generallyconsidered a useful measure of incubator value.An emphasis on job creation also contradictsthe advice of many investors who are acutelyaware of the need to control spending byinvestee rms, which often means delayingrecruitment. Further work is needed todevelop appropriate performance indicators forincubation.

In practice, incubation can lead to several

outcomes for new ventures. Incubation canimpact new ventures through modifying oraccelerating the entrepreneurial process ofbusiness development. But while incubatorshave been associated with businessacceleration of incubatees, this same processcan lead to life support which extends thetime to business failure. A period of high riskcan confront incubatees when leaving thesupport of an incubator.

Selecting rms with potential for high growthis an uncertain process. A portfolio approach

mitigates the risk associated with relying on theperformance of a single rm. Across a portfolioof incubator tenants around 23 per centidentify the incubator as important to businessperformance.3 Over 60 per cent identify theincubator as critical, while just under 17 percent regard the incubator as unimportant to

performance.

Incubators inuence new rms by:

Lending credibility through association, andthrough shared (and therefore affordable)access to professional facilities and anidentiable and exible incubation space.

Offering business support and coachingwhich are often subsidised e.g. strategicinsights, market research etc.

Providing access to additional resources andtalent e.g. nance, legal help.

The incubator draws on its own staff, externalconsultants, and its existing entrepreneurialsupport network to provide business support.Peer-to-peer networking is also encouraged.

Matching incubator services to the needs of

rmsis important. New venture activity andbusiness support needs vary between regions,industries, prior entrepreneurial experience and

so on.

Incubators with links to universities areassociated with technology rms with highergrowth potential, but not all universities havean entrepreneurial culture or are surrounded bya supportive business environment. In additionto technology and facilities, people are a maincontribution of universities to entrepreneurialactivity.

Rather than cater to all rms, most businessincubators have a selection/screening processto target a particular group of rms. Thisscreening process is imperfect, but can beimproved through the use of multiple screeningdimensions.4 Nonetheless a selection processcan only be imposed if the incubator can affordto turn away potential tenants.

Tenants seem to become dissatised withincubator support when the incubationprogramme is predetermined rather thanre-evaluated depending on the changingneeds of tenants. The entrepreneurial support

mechanisms also uctuate, with incubatorsable to offer some continuity.

2. Membership schemesincentivise a broad view ofincubation to include as manymembers as possible.

3. CSES (2002) Benchmarkingof Business Incubators.Sevenoaks: Centre forStrategy and EvaluationServices.

4. Aerts, K., Matthyssens, P.et al. (2007) Critical roleand screening practices ofEuropean business incubators.Technovation. 27: 254-267.


7/54

As incubators become more embedded in aregion they tend to become more specialised. Aword of caution while many try and emulateincubation strategies from Boston, SouthernCalifornia (US) or Cambridge (UK), theseregions are also considered atypical and likenedto regional incubators owing to the amount

and maturity of entrepreneurial activity andinfrastructure.

Even incubators with similar objectives can

have different business models. Businessmodels have changed over time. Since 2005there has been a reported increase in thecost per job created each year in businessincubation in Europe. Some incubators nowoffer equity nance, and some equity investorsoffer incubation, with an unclear distinctionbetween both. The challenge for incubatorsand their funding bodies is to capture some of

the value created for incubatees. Generatingrevenue from services when clients are resourceconstrained is often not possible withoutsubsidies from public bodies. Corporate fundedincubators typically require a strong strategict of incubatees with the corporation, which isnot appropriate for all ventures. Incubators withmixed funding may encounter principal-agentproblems as they attempt to meet multipleobjectives.

Capturing value through taking equity in

clients introduces delays in revenue and cancause the incubator to behave more like anequity investor by prioritising short-termnancial returns rather than longer-termperformance. The literature offers little insighton whether incubators could generate betterreturns for early-stage investments thanpure equity investors. Already early-stageinvestments are associated with poor returnsin Europe, especially compared to the US.Further research is needed to understand thestrengths and weaknesses of business modelsfor different contexts.

In summary, the evidence we have reviewedindicates business incubation is a valuabletool as part of an entrepreneurial supportinfrastructure. Incubators deliver the mostvalue when able to respond and adapt to theneeds of new ventures. We realise some of ourconclusions regarding how business incubationshould be monitored challenge some existingnorms in this domain. However, the lack ofcomparability between studies demonstrateshow important it is to improve the quality of

metrics. Even so any measure of incubationis likely to be incomplete. The impact ofincubation on incubatees should extend

beyond the incubation period and incubatorenvironment, though measuring this impactcould become onerous and time consuming.

While we recognise the variety of businessmodels used and the continuing evolutionof the industry, we nevertheless conclude

that further research is required for thefundamentals of incubation models a topiclargely neglected in the extant literature tobe properly understood.


8/54

Figure 2: Overview of themes and research questions incubator-incubation literature

Incubator

development studies

(1984-1987)

Defnitions

Taxonomies

Policy prescriptions

What is an incubator?

How do we develop anincubator?

What life cycle modelcan be extracted?

Formal analysis ofbusiness incubators

Incubator

conguration studies

(1987-1990)

Conceptual rameworks

Incubatee selection

What are the criticalsuccess factors forincubators-incubation?

How does theincubators-incubationconcept work inpractice?

How do incubatorsselects incubatees?

Incubator

development studies

(1987-1988)

New venture

development

Impact o planning on

development

What is the processof new venturedevelopment in anincubator context?

What is the role ofplanning and thebusiness incubatormanager?

Incubator-incubation

impact studies

(1990-1999)

Levels and units o

analysis

Outcomes and measures

o success

Do incubators achievewhat their stakeholdersassert they do?

How can businessincubation programmesoutcomes be evaluated?

Have business incubatorsimpacted new venturesurvival rates, jobcreation rates, industrialinnovation rates?

What are the economicand scal impacts of anincubator?

Theorizing about

incubator-incubation

(1996-2000)

Explicit and implicit

use o ormal theories

(transaction cost

economics, network

theory, entrepreneurship,

economic

development through

entrepreneurship)

What is the signicanceof relationships andhow do they inuenceentrepreneurship?

What are the criticalfactors to success e.g.,settings, networks,founder characteristics,group membership,coproduction value and

creation process?

Source: Hackett and Dilts 2004b.

Since the rst recognised incubator establishedin Batavia, New York in 1959, there was a

slow diffusion of incubator programmes inthe 1960s/70s. Incubator diffusion increasedsignicantly in the 1980s/90s (Figure 3) inconjunction with increased attention and clarityon the commercialisation process of research5(Hackett and Dilts 2004b). The establishmentof industry organisations like the US NationalBusiness Incubation Association (NBIAestablished in 1984) and the United Kingdom

Business Incubation (UKBI established in1998), led to increased interest in how to

measure and report incubator performance.Academic contributions soon grew andhave continued to evolve to this day (for asummary of contributions see Hackett and Dilts2004b and Phan, Siegel et al. 2005) (Table1). While many for-prot incubators closedwhen the dotcom bubble burst, incubationhas nonetheless prevailed as part of a widerinnovation system (Hackett and Dilts 2004b).6

Part 3: Research background

5. For example the Bayh-DoleAct 1980 decreased theuncertainty associated withcommercialising federallyfunded basic researchand intellectual propertyrights protection becomeincreasingly recognised(Hackett and Dilts 2004b).

6. For an example ofcommentary on internetincubators, see: http://www.strategy-business.com/article/11071?gko=a2013


9/54

Since the 1980s business incubators havebecome a popular policy instrument tofoster entrepreneurship, innovation, and

regional development (OECD 1997; CSES2002). Incubators are not the only potentialinstrument for achieving these goals, nor canthe presence of an incubator alone securethem (Phan, Siegel et al. 2005; Connell andProbert 2010; Hussler, Picard et al. 2010).Nonetheless the continued growth in thenumber of business incubators worldwidedemonstrates their perceived value. With theproliferation of business incubation activitiesover the last 50 years has come diversicationin the terminology used and type of incubationactivities offered.

3.1 Objectives of incubation

Two main rationales for incubation haveemerged. The rst view regards incubationas a way of addressing market failures, whichlimit the ability of small high-tech start-ups toovercome uncertainty and obstacles associatedwith the early stages of rm development(OECD 1997; Phan, Siegel et al. 2005). Market

failures stem from the relatively high costsand risks associated with providing supportto high-tech start-up companies. Privatesector institutions are unwilling to absorb the

costs and risks if the commercial value of thetechnology being exploited is too uncertain asis often the case with early-stage ventures. The

second view regards incubation as a catalystto accelerate the entrepreneurial processsystematically, thereby institutionalising thesupport of ventures with potential for highgrowth (Hansen, Chesbrough et al. 2000). Inpractice incubation has been associated witha variety of objectives (Allen and McCluskey1990) (Table 1). Very often these objectivesrelate to the specic business environment(nationally and regionally) in which theincubator is located.

The primary incubator function has beendescribed as increasing the chances of anincubatee rm surviving its formative years(Allen and Rahman 1985). Theoreticallythe incubator can also impact an individualtenant through improving its growth path,as illustrated in Figure 4. This impact can lastbeyond the incubation period. As a resultincubation can full many of the objectivesdescribed in Table 1, through enhancing growthin the productivity and employment of itstenants both during and after the incubationprocess, which in turn has an impact on the

wider business environment. The rest of theliterature review explores this idea in more detailto properly understand the impact of incubation.

Figure 3: Growth of the worldwide incubator industry

Source: Barrow 2001.

1,500

2,000

1,000

500

4,000

3,500

3,000

2,500

0

1st wave 2nd wave

1980 1985 1990 1995 2000 2005


10/54

0

Primary objective

Real estate appreciation

Sell proprietary services to tenant

Job creation

Positive statement of entrepreneurial potential

Faculty-Industry collaboration

Commercialise university research

Capitalise investment opportunity

Secondary objective

Create opportunity for technology transfer

Create investment opportunity

Generate sustainable income for the organisation

Diversify economic base

Bolster tax base

Complement existing programmes

Utilise vacant facilities

Strengthen service and instructional mission

Capitalist investment opportunity

Create good will between institution andcommunity

Product development

Table 1: Incubation objectives

Figure 4: Theoretical impact of an incubator on the irregular growth path of an individualtenant

Source: Allen and McCluskey 1990.

Source: Authors own interpretation.

Growth

Incuba

tionp

eriod

Time



11/54

1

3.2 Typologies of incubators

Even incubators with similar objectives mayhave different typologies (Bergek and Norrman2008). The literature remains unclear on how toorganise typologies of incubators (Hackett andDilts 2004b; Phan, Siegel et al. 2005). Terms like

research park, technology innovation centre,science park, business incubator, acceleratorhave become interchangeable (Hackett and Dilts2004b; Phan, Siegel et al. 2005).

This literature review focuses on studiesexploring incubators that include physicalspace7 and business support interventionsgeared towards high-growth rms. We adoptthe denition from Hackett and Dilts (2004b)where business incubation is dened as:

...a shared ofce-space acility that seeks

to provide its incubatees with a strategic,value-adding intervention system o

monitoring and business assistance.

This contrasts earlier denitions of incubationwhere survival of tenants is emphasised (Allenand Rahman 1985). Survivability is a limitedmeasure of how a rm is performing, and is anecessary but insufcient condition for success.If success is dened as the achievement o

something desired; planned or attempted

(Oxford English Dictionary 2010) thenentrepreneurs in pursuit of high growth mayconsider survival, without growth, a failure.

It is recognised that business incubators cancreate value for a variety of stakeholders butthey also depend on a variety of stakeholdersto have a viable business model. Hackettand Dilts identify value creation betweenincubatees, incubators and the community8(2004b). Building on this we have identiedperspectives in the incubation literatureclustered around several thematic axes (Figure

5). These axes have been organised arounddifferent levels of analysis and are discussedin more depth throughout the report. Figure 5

Figure 5: Illustration of the range of business incubation perspectives through referenceto thematic axes representing different levels of analysis9

Source: Authors own work.

7. There are mixed views onwhether virtual incubators,i.e. incubators without walls,are incubators at all (Bearse1998 cited in Hackett andDilts 2004). There has beena rise in a variety of virtualincubators and accelerationprogrammes as incubatoractivities have diversied. This

study focuses on incubatorswith physical space, a spacewhich has been likened to aclubhouse. While space isnecessary but non-sufcientfor these incubators, it alsohas convening power whichis enhanced when the spaceis designed for businessincubation. Anecdotally it hasbeen suggested that physicalspace encourages more face-to-face interactions, whichbuild a greater level of trustthan can be achieved online.This trust encourages moremeaningful interactions andexchanges of knowledge,and being part of the space

increases the frequency ofsuch interactions.

8. The term community refersto the business environmentsurrounding the incubator.

9. For further information aboutthese axes, please referto the rest of this report.Spin-outs refer to businessesofcially spinning out of theuniversity environment, whereas start-ups are any typeof new business. A start-upmay refer to an unofcialbusiness which spun-out ofthe university, which is whythe axis is presented as acontinuum.

Technological level

Incubatees

Low High

Incubatee originSpin outs Start ups

Selection strategy of incubatees

Incubator services

Weak Strong

Value adding (in addition to real estate)Low High

Funding

Incubator as an organisation

Private Public

Life cycleNew Embedded

Supportive to target ventures?

Incubatees

Low High


12/54

2

is designed to illustrate the potential range ofbusiness incubation perspectives but may notbe exhaustive.

In attempts to engage with those interestedin incubation, the term has been looselyapplied to various activities at different

units of analysis. Increasingly organisationsinvolved with the support of entrepreneurialactivity risk being referred to as incubators,with some extending the term to unusuallyentrepreneurial regions (Phan, Siegel et al.2005). Some venture capitalists who offerhigher than average business support activitieshave also been termed incubators, particularlyin the US. This confusion between investors,professional services rms offering businesssupport, and incubators has prompted somecomparative studies (Hsu 2007; Aaboen

2009). We have built on these to producea top-line comparison between incubation,venture capital and professional services rmsto identify whether incubation is unique (Table2). For example, though not as intensive as forventure capitalists, the implementation of aselection process for tenants seems important

in incubation.

10

In addition to businessassistance, the aggregation and interactionof incubatees co-located inside the incubatorhas been identied as unique to incubation(Hackett and Dilts 2004b). Table 2 supportsthis and suggests the distinctive features ofincubation are a very mixed revenue stream,strong encouragement of peer-to-peernetworking, addressing multiple needs of newventures without prioritising just one, andoffering continual exposure to the incubationenvironment and services.

Deal ow/clients

Main revenue

stream

Primarily addresses

what market need?

Peer-to-peer

networking?

Time scales

Target rms

Venture capital

Wide search processes,sometimes regional or industryspecic. Before securing a clientthere is intensive due diligence

Returns on investment inventures

Need for equity nance to fundhigh growth

Usually limited

Usually seek an exit 3-5years after investment, butinteraction with ventures is

episodic

Typically addresses a narrowerrange of rms than incubators,and at a later stage

Professional services rms

Often bid for projects

Billable hours

Address a knowledge gap inclients

Usually restricted

Depends on a project byproject basis, but typicallymonths not years

Broader range of rms, nottypically restricted to newventures

Table 2: Identifying the uniqueness of incubation

Source: Authors own analysis with reference to Aaboen 2009, Hsu 2007, Hackett and Dilts 2004.

Incubation

Receive applications from ventures that aresubject to a selection process e.g. incubatorbranding encourages self-selection or selectioncriteria is imposed on potential tenants

Mixed revenue between rental income and otherpublic and private sources (e.g. 1)

Access to space, knowledge, resources, via staff,programmes and networks

Actively encouraged and facilitated through ashared space (typically physical space but alsopossible in an online space)

Typically 3-5 years (anchor tenants are oftenlonger) with a mix of episodic and continuousinterventions but continual exposure to

incubation environment

Typically addresses a broader range of rms thaninvestors, and usually at an earlier stage thanVenture Capitalists

10. Professional services rmshave some selection processfor clients.


13/54

3

There is no standard methodology formeasuring incubator performance, which makes

comparisons between studies challenging(Phan, Siegel et al. 2005) (Appendix A). Manyincubators are non-prot which renders theusual economic analysis challenging. Eventhose incubators identied as private oftenhave public support for programmes they run.It is difcult to distinguish between rm growththat would occur in the absence of incubation,and additional growth as a result of incubation.To collect data many studies survey incubatormanagers as a central point of contact, this isuseful but limited as they cannot accurately

represent the views of incubatees. A summaryof approaches is offered in Table 3, butfew studies have uncovered meaningfulcategorisation processes linking relevantfactors to specic contexts. As a result manyndings have limited generalisability.

4.1 Incubation industry reports

Associations of the business incubator industryfrequently assess the impact of business

incubators and offer estimates of aggregateperformance. Organisations such as the US

National Business Incubation Association aremembership-based, which incentivises theinclusion of as many members as is reasonablypossible, as is evident from the lack ofscreening of new members. As a result it isdifcult to build a high degree of condencearound the homogeneity of their data sets.

Most incubators remain either wholly orpartly publicly funded. In the competition toattract public funds many incubators needto constantly demonstrate success, which

can lead to over-reporting successes andunder-reporting failures especially whenself-reporting (Hackett and Dilts 2004b).We propose a cautious view of the followingincubator industry data:

In 2005 alone, the NBIA estimates that NorthAmerican incubators (~1,100) assisted morethan 27,000 start-up companies that providedfull-time employment for more than 100,000workers and generated annual revenue of morethan $17 billion (based on extrapolations fromsurvey data) (Knopp 2007).

Part 4: Evidence on the impact of business incubation

Approach

Control-groupconcept

Benchmarking

In situ assessments,before and after

Method

Pairing rms,on and off site

Surveys, categorisation,

interviews, self-reporting

Surveys, case studies

Examples

(Lindelof and Lofsten 2002; Siegel, Westhead et al.2003b; Dettwiler, Lindelof et al. 2006; Amezcua 2010)

(CSES 2002; Knopp 2007; UKBI 2009)11

(Grimaldi and Grandi 2005; Bergek and Norrman2008; Patton, Warren et al. 2009)

Table 3: Review of research methodologies used to assess incubator performance

11. While these studies attemptto benchmark incubatorperformance, their mainmethodology is an in situassessment.


14/54

4

Business incubators in the EU which nownumber around 900 make a signicantcontribution to job and wealth creation.Some 40,000 new (net) jobs are generatedeach year by incubators (CSES 2002).

The UK has a well established network of

approximately 300 business incubators thatsupport over 12,000 high-growth technologybusinesses in sectors such as biomedical, ITand the creative industries (UKBI, 2010).12

While these industry reports are likely tohave an optimistic view in promotion of theirindustries, they indicate strong activity linkedto business incubation. The range reportedis between 25-40 supported businesses perincubator, and between 44-91 jobs created peryear per incubator.13 But these gures typicallyinclude a mix of technology and other types of

incubators. Job creation remains a limited butpopular measure used to evaluate incubatorperformance (CSES 2002; Frontline 2002;SQWConsulting 2008).14 Yet new ventures willoften try to reduce their xed costs as theyoperate in conditions of uncertainty. Ventureinvestors are acutely aware of the need tocontrol spending by investee rms, which oftenmeans in practice delaying recruitment of full-time employees (FTEs) as long as possible andinstead preferring the use of exible contractworkers and consultants. This can lead to

conicting goals as incubators try to satisfythe needs of public bodies through supportingjob creation, but also the needs of investors bydiscouraging incubatees taking on additionalrisk through recruiting FTEs.

4.2 Quantitative academic studies

Despite the growth in literature on incubation,few studies have applied a robust evaluativeapproach to assessing the economiccontributions of incubators. Many quantitativeacademic studies attempting to evaluate theimpact of incubators on populations of rmshave more conservative results than industrystudies, and often contradictory ndings.Furthermore some of these studies focuson science parks as opposed to businessincubators as few studies have access tomeaningful quantitative datasets for businessincubation alone, which makes comparisonschallenging. For example an analysis of alongitudinal UK data set of on- and off-science

park matched rms, including two surveys in1986 and 1992, shows a higher survival rateamong rms on parks with a higher education

institute (72 per cent) than without (53 percent), insignicant difference between closurerates, and growth in employment concentratedin the hands of a few businesses, whilemean employment growth rates were similar(Siegel, Westhead et al. 2003a). In our reviewof studies using more recent data sets, we

found few contributions offering additionalinsights. Chens (2009) study of Taiwaneseincubators found no direct effect on newventure performance as a result of incubation,whereas Rothaermel and Thursby (2005)showed incubated rms were signicantly lesslikely to experience outright failure. Lindelofand Lofsten (2002) discovered no difference inprotability between on- and off-park rms,but the off-park sample had signicantly lowergrowth in employment and sales turnover.Making sense of such ndings requiresscrutiny of the research designs employed and

their limitations (Appendix B). For exampleRothaermel and Thursby (2005) used a singleincubator study which is a useful exploratorycontribution to incubation research but withlimits to generalisability.

This set of academic studies highlights thedifculty in answering what at rst looks likea straightforward question do incubatorshave a positive impact? As the outcomes ofincubation may take many years to becomeapparent, as a company develops its markets

and scales its production, success varies fromwhether incubated ventures survive longer orhave signicant growth whilst being incubated.Many different approaches have been taken insignicantly different time periods and contexts(Appendix A & B). Whist there are no highlynegative outcomes, the positive outcomesare based around survival (in the case ofRothaermel and Thursby 2005) or higheremployment growth (Lindelof and Lofsten2002). Aside from direct measures of successfor incubated rms, the empirical evidencewould suggest that incubatees who interactwith the incubator (both in terms of othercompanies and support staff) have strongerlearning (Scillitoe and Chakrabarti 2010), whileincubators who screen against a balanced setof indicators will have lower failure rates (Aerts,Matthyssens et al. 2007).

Taken together the studies are indicative ofthe approaches that may work, however giventhe relatively small number of studies and thelack of comparability between studies anyconclusions should be treated as indicative at

best.

12. See:http://www.publications.parliament.uk/pa/cm201011/cmselect/cmsctech/writev/innovation/m16.htm

13. The US EPA recently

evaluated the cost of jobsin various industries, andconcluded that businessincubation was the mostcost effective job creationmechanism. Again, werecommend examiningthe methodology (source:http://www.eda.gov/PDF/EDACons ImpactStudyVolume1FINAL.pdf)

14. Examples drawn from theassessment of businessincubation by public bodiesinclude requirements tomeasure core outputs.Drawing on real examples,these are often identied

as number of jobs created,number of jobs safeguarded,number of businessessupported/assisted toimprove performance. Eachoutput claim has to havefull documentation, whichis challenging with newventures that very rarelyhave automated systemsto generate the requiredevidence. One such sourceof funding for incubationwithdrew some funding inorder to pay a part-timemanager to ensure the papertrail was complete.


15/54

5

4.3 Impact of incubation on the widerbusiness environment

Incubators are capable of extending theirservices beyond their community of incubatees,and incubatees extend their business beyondthe walls of the incubator. When combined

with a perspective of the incubation periodgenerally being shorter than the life cycleof a rm, the impact of an incubator on thewider business environment is typically greaterthan any common measure recognises. Whilethese business incubation effects can createconsiderable value in the wider businessenvironment, they can be challenging tomonitor. Wider impacts of business incubationbeyond their direct affect on tenants include:

Incubation outreach The NBIA noted in2006 a slight upward trend in the services

being offered by incubators to entrepreneurswho do not reside within the incubator. Theycan extend their reach through assistingnascent entrepreneurs pre-incubation andsupporting post-incubation entrepreneurs(Knopp 2007). UKBI recently estimated that60 per cent of incubators operate what theyterm as outreach services that support andadvise companies outside the incubator.15

Indirect effects For example displacementof non-incubated rms with incubated

rms, or additional jobs and wealthgeneration from supplying goods andservices to incubator and tenants, orincreased local spending from employeeincome of incubated rms (CSES 2002). AEuropean study estimated a ratio of 1:1.5 fordirect:indirect job creation from incubatees(CSES 2002). Another study suggests moreconservative indirect effects with multipliersbetween 0.48 and 0.84 for the number ofindirect jobs created from direct jobs fromtwo incubators (Markley and McNamara1996).

Entrepreneurial learning New venturefailure is often documented as the endof an entrepreneurial process, but itfrequently spawns other opportunitieswhich entrepreneurs are in a better positionto pursue after learning from failure16 Iasked about the chie product o some o

these frms, one might reply entrepreneurs.

(Cooper, 1971 p.2 cited in Garnsey andHeffernan 2005b). For example, even thoughAcorn Computers no longer exists, it was a

fertile ground for entrepreneurial learning,and its founders and employees used their

experience to establish many new rms(Figure 6).

In addition to the indirect effects described,an incubator can become a representativeof entrepreneurs, offering a single point ofcontact for those wishing to engage with

new ventures. With the rapid uctuationstypical in a population of new ventures,this can be a valuable source of informationon entrepreneurial activity which can becommunicated more widely, for example topublic bodies and government.

4.4 Absolute measures of incubation areimpractical, but performance indicatorsare useful

While monitoring incubator activity is generallyconsidered useful by incubators and theirstakeholders, it can also become cumbersomeand erode the ability of the incubator toperform its core functions. One study identiedthat incubator managers were less effectivewhen distracted from their core activities byexcess monitoring or the need to secure fundsfor the business incubator (Rice 2002). The2006 NBIA survey team believe the lengthof the survey17 was to blame for their lowestresponse rate since their surveys began.

An alternative approach was taken by theSwedish VINNKUBATOR programme (nowInnovationsBron18) where they asked FokusAnalysis to develop an online assessmenttool in collaboration with incubators. The toolwas to be of use to incubators for monitoringtheir own performance, but also enabled acentralised collection of data for review bypublic bodies. This approach seems to havebeen well received, and data is regularlycollected. Nonetheless surveys and monitoringare very unlikely to measure all activities oroutputs of business incubators, and attemptsto would likely be cumbersome and timeconsuming. The literature advocates the useof longitudinal data to explore incubation, sopractical data19 collection must be considereda priority in addition to collecting appropriatedata to indicate incubator trends.

15. See: http://www.ukbi.co.uk/about-ukbi/business-incubation.aspx

16. One documented exampleof this is Acorn Computerswhich was founded in 1979in Cambridge, UK. Afterexplosive growth in demandfor Acorn microcomputersthere was a sudden slump inconsumer demand in 1984.The company only survivedits over-commitment tosuppliers through acquisitionby Olivetti. Acorn waswound up in 1999, withOlivetti benetting fromshares in ARM. WhetherAcorn is judged in itsown right as a success orfailure is only part of thestory as it was a valuableenvironment in which manylocal entrepreneurs and

managers gained experience,which helped produce otherventures (Garnsey andHeffernan 2005b).

17. The survey included 31multi-part questionsrequiring over 200 answers.

18. Information on the FokusAnalysis can be found atwww.innovationsbron.se

19. For example see Hackett andDilts 2004.


16/54

6

Figure 6: New rms started by founders and employees of Acorn Computers

Source: Garnsey and Heffernan 2005b.

2006

Olivetti Research, 1986 (AT&T)

ATM, 1993 VIRATA

SynGenix, 1992

IQ Bio, 1981

Orbis,

1978

Clearswift (Net-Tel Computer Systems), 1982

GIS, 1982

ANT, 1993

STNC, 1993

nCipher, 1996

Xemplar Education, 1996

Commtag, 2000

Icero Semiconductors, 2003

Element 14, 1999

Pogo Mobile Solutions, 2002RealVNC, 2002

Ubisense (Ubiquitous Systems), 2002

Level 5 Networks (Cambridge

Internetworking), 2002

Cambridge Broadband, 2000

(Cambridge Network Ltd, 1998)

Vocals, 1992

Ed Inc (with AT&T)

IXI Ltd

ABC, 1986

Harlequin Ltd, 1986

Qudos Technology Ltd, 1985

Advanced Displays

Electronic Share Information Ltd, 1993

NetProducts Ltd, 1996

NetChannel Ltd, 1996

Amadeus Capital

Partners, 1997

E*Trade UK, 1996

Adaptive Broadband, 1998

IPV (Telemedia Systems), 1995

ARM, 1990

2000

1995

1990

1985

Pre 1980


17/54

7

As described earlier, regardless of anincubators specied objective, the incubators

main lever with which to achieve its objective isthrough its impact on clients i.e. new ventures.We explore this direct impact on new venturesin more detail.

5.1 New venture outcomes as a result ofbusiness incubation

Measuring the performance of new venturesremains a challenge to researchers worldwide.

While established businesses are typicallyassessed in terms of share-value or grossprot, new ventures rarely have either.Therefore assessing the impact of incubatorsis complicated by the lack of consensus onhow to measure the performance of early-stage rms. Furthermore, the incubationperiod is typically shorter than the life cycle ofa rm, so assessing rm performance duringthe incubation period misses longer-termeffects. Due to the irregularity of most rmgrowth paths (Garnsey and Heffernan 2005a),estimating rm performance had incubationbeen absent proves challenging.

Hackett and Dilts propose an application of realoptions to incubation in an attempt to offer arobust way of assessing incubator value andmanaging capital investments (2004a). A realoptions approach, originally used in corporatenance, applies option valuation techniquesto capital budgeting decisions, so creatingoptions for when to make, abandon, expandor contract a capital investment. Five optionsare described under initial outcomes (Table

4). That practice is less pure than theory isimplicitly acknowledged by Hackett and Dilts,However, acilitating the survival o incubatees

or containing the cost o ailure o the options

to the sunk cost o creating the option minus

any remaining option value, and reportingthese successes, can result in the renewal o

annual operating subsidies, a very important

upside without which many incubators would

close.(2004a p.51 emphasis added). Inshort, real options theory provides valuabletheoretical insight into incubation practice.But since incubatee options cannot be pricedaccurately until they are realised or expire(too many uncertainties prevail for accuratequantication), options theory is insightfulrather than universally true. David Gill, manager

of St Johns Innovation Centre, offers thoughtson the application of real options to anoperational incubation environment, showingthe value and limits of its application (Box 1).

As recognised in the real options framework,incubation doesnt always lead to betteroutcomes for incubatees. While supportersof incubation suggest the process can helpshield incubatees from competitive forcesof the external environment and increasethe likelihood of short-term survival, otherscontend that this same process can weakena rms ability to compete and survive whengraduating out of the incubator (Amezcua2010). In evolutionary theory a rm will fail ifit develops routines and competencies that aremisaligned to the competitive selection regimeof the business environment (Aldrich 1999).A recent US study has found incubated rmsoutperform their peers in terms of employmentand sales growth, but fail sooner (Amezcua2010). Few studies explore post-incubatorperformance, and yet graduation is easy,post-graduation survival may not be (Schwartz

2010).20 A study of German incubators founda period of high risk confronts graduateswithin their rst three years after graduation

Part 5: Understanding the impact of business incubation

on new ventures

20. The shock came afterleaving the incubator inrealising what actuallyneeded to be done to existin your own premises. (UKBI2009 p.40).


18/54

8

Antecedents

Pre-ventureinitiativeactivities

Communitysupport forentrepreneurship

Exogenousconduct of basicresearch

Events increasingindividual

entrepreneurialorientation

Incubatorfeasibility study

Inputs

Entrepreneurs

Enablingtechnologies/innovations(includingincubator)

Criticaltechnologies/innovations

Outputs

Incubatedcompanies

Initial outcomes

1. Incubatee is survivingand growingprotably

2. Incubatee is survivingand growing but notyet protable

3. Incubatee is survivingbut not growing andnot protable

4. Incubatee operationsterminated while

still in the incubator:losses minimised

5. Incubatee operationsterminated inincubator: largelosses

Intermediate

outcomes

Viable/becomingviable companies

Dead dyingcompanies

Table 4: Real options framework

Activities

Incubation:New venturedevelopment+ new productdevelopment+ selection +monitoringand businessassistance+ resourcemunicence

Long term

outcomes

IncreasedOrganisationPopulation Churn

Source: Hackett and Dilts 2004a.

Figure 7: Growth impact of the misalignment of the incubation and business environment(representation of life-support e.g. Schwartz 2009)

Growth

Time


Incubation

period


19/54

9

Figure 8: Illustration of potential growth scenarios for incubatees as a result of incubation

21. New organisations usuallyhave limited credibility thatrestricts the acquisitionof resources. Liabilities ofnewness are well-recognisedfor founding organisations inthe literature (Stinchcombe1965).

Growth

Time

Long term value added Original growth path Life support Temporary value

Incubation

pe

riod

where around 20 per cent of graduates do notsurvive (Schwartz 2009). They suggest this

originates from extended life support of someincubatees who should have perhaps beenmore closely monitored. This life-support ofincubatees is represented in Figure 7. Beyondthe rst three years after graduation there isa high probability of durable establishment ofthe graduate (Schwartz 2009).

Another risk of incubation is that the impactis very temporary. The outcomes of incubationdiscussed are illustrated in Figure 8. The impactof an incubator overall will depend on theportfolio of incubatees and the impact of theincubator across the portfolio. How much anincubator can impact new ventures depends onthe incubation tools available, in addition tocharacteristics of the new venture.

5.2 How incubators inuence outcomesin new venture performance

It is useful to distinguish between the potentialbenets to incubator tenants, and how these

benets are delivered by the incubator. For theformer it has been suggested that incubatorsprovide benets to incubatees along four broad

dimensions: (1) development of credibility,21(2) shortening of the entrepreneurial learning

curve, (3) quicker solution of problems, and (4)access to an entrepreneurial network (Smilor1987). To examine this further we sought inputfrom research in entrepreneurship which drawson more established work on this subject thanin the business incubation literature.

A conceptual framework for the businessmodel, or entrepreneurial process, of new rmshas been developed by Garnsey (Stam andGarnsey 2005; Dee 2008) (Figure 9). This helpselucidate the ways an incubator could inuencethe trajectory of rm development. Every rmcustomises the model taking account of theirown resource base and perceived opportunities.New rms tend to be more focused on businessideas and gaining the resources needed tobuild a productive and commercial base, whilemore established rms focus on value creationand capture. Incubation can impact variousaspects of the entrepreneurial process, fromstrategic input to the business model, tomodifying or accelerating the entrepreneurialprocess through access to resource providers,entrepreneurial learning from peers, access

to customers, advice on intellectual propertyrights to improve value capture, etc.


20/54

0

Box 1: Real options from the perspective of an incubator manager DavidGill, St Johns Innovation Centre, Cambridge UK

The Hackett and Dilts real options framework lists ve potential outcomes to the incubationprocess and suggest that outcome 4 (operations of the incubatee are terminated, withlosses thereby minimized) is a success and outcome 3 (an incubatee becomes part of the

living dead, to borrow a term from venture capital) a failure, is insightful in describingsome of the ltering functions of incubators (and would be true of a venture fund orincubator operating on the venture investment model) but may not apply in all incubationcircumstances. That practice is less pure than theory is implicitly acknowledged by Hackettand Dilts.

Most subsidy-providers would not operate within the options framework and wouldinstinctively prefer an option 3 outcome to an option 4 outcome. Subsidy-providers arelikely to be public authorities who will use proxy measures of success, such as number ofjobs created, or investment attracted, or survival rates n number of years after the companyleft the incubator. Funding sources generally rely on intermediate outcomes at least as muchas they use the hard measures of real growth and protability.

Part of the reason is down to timeframes. While the authors are right to say that incubatorsare not the all-powerful innovation hatcheries capable of incubating and taking publicinfnitely scalable, dot-com e-business start-upsless than a year after entering theincubator, option theory does not take fully into account the potentially elastic timescalesinvolved where incubation is concerned a crucial difference when incubation (otherthan short-life accelerator programmes) is contrasted with a ten-year xed-life limitedpartnership, common in venture capital.

Tenants may quite legitimately linger on in a physical incubator, often justiably takinglonger to bring projects to fruition than originally envisaged. Their best ideas can be theones that grow out of their original proposal and which need to go through numerousiterations before becoming a killer app.

Another fundamental difference between the VC option model (relatively purist) andthe incubator option model (much more pragmatic) is that VC funds have partners orshareholders, whereas incubators usually have multiple stakeholders. VCs operate a prettyclear shareholder business model (their shareholders are usually limited partners for taxpurposes) but almost all incubators are much more reliant on having to please a range ofstakeholders even where an owner applies commercial targets.

A venture investor is 100 per cent reliant on getting the portfolio right managing realoptions. The VCs success is ultimately dependent on selling its cashed-in options for manymultiples of the original option price, and wise VCs spend as little time as possible on theliving dead. But the incubator manager has several sources of income rent, grants, publicor private consulting, conferencing and other services, maybe share options as well andso also numerous constituencies to please, whose interests may not always be obviouslyreconcilable.

A shrewd incubator manager will team and ladle different sources of income to cross-subsidise different activities, something pure real-options theory does not allow for. This islegitimate in terms of fullling a mandate of supporting growth rms because of the highdegree of uncertainty involved in taking on a new client. Success depends on the ability andexperience of the incubator team, who often act with limited information and so must relyon judgment as much as formulae. The black box cannot be entirely dispensed with.

The authors acknowledge that resourcesper se are not enough if there is no accompanyingselection process. However, selection takes place at multiple intervals during (also before

and after) the relationship and not just up front. Selection is not just about dealing withexisting tenants either, it also involves long-term relationships with individuals who move


21/54

1

Figure 9: The entrepreneurial process

Productive/

commercial

base

Co-producers

Customers

CompetitorsFounders

Resource providers

e.g. enabled

by financiers

Value creationBusiness idea

Value

capture/

Asset base

out but may move back in, their advisers and investors something very hard to put a priceon, and so value as an option.

In short, real options theory provides valuable theoretical insight into incubation practice. Ithelps provide a framework to deal with the uncertainties inherent in incubation. But sinceincubatee options cannot be priced accurately until they are realised or expire (too many

uncertainties prevail for accurate quantication), options theory is insightful rather thanuniversally true.

The implication for policy purposes is not that all incubators should be run as short-termaccelerators rewarded with equity kickers like a VC, but that subsidy providers would benetfrom being more sophisticated in measuring success and less reliant on proxy measures suchas numbers of jobs created in a short space of time.

David Gill (St Johns Innovation Centre and IfM) started his career in investment bankingwith Chase Investment Bank, moving to Greenwell Montagu in 1988, from where he joinedthe corporate nance department of Midland Bank plc, specialising in advising smallergrowth companies. After the acquisition of Midland by HSBC he became Head of theInnovation & Technology Unit for the UK bank. He was a director of ETCapital, a specialist

technology seed-fund investor, before being appointed MD of SJIC. In 2004-5 he was aSloan Fellow at the Stanford Graduate School of Business. He was an undergraduate atMagdalene College and was called to the Bar by the Middle Temple.


22/54

2

As a result of insights into the entrepreneurialprocess, we contrast Smilor (1987) and suggestthe key benets to entrepreneurs of beingin an incubation environment are to modifyor accelerate the entrepreneurial process.This is achieved through the development of

credibility, shortening of the entrepreneuriallearning curve, quicker solution of problemsand access to an entrepreneurial network.

As we are focused on incubation for rmswith potential for high growth, we now offera perspective on these kinds of rms beforeturning to how incubators deliver value.

5.2.1 High-growth rms and a portfolio

approach

A thorough explanation of high-growth rms

(HGFs) goes beyond the scope of this work.Drawing on our own and others work wesuggest high-growth rms share the followingcharacteristics:

There remains a lack of consensus on howto measure the performance of early-stagerms, especially pre-revenue (Garnsey andHeffernan 2005a).

Firms increasingly source knowledge from avariety of sources, not just the science base(Huggins, Izushi et al. 2010).

Firms with high-growth episodes aredisproportionately important to theemergence of industries, job creation andeconomic wealth (Acs, Parsons et al. 2008;Garnsey and Mohr 2010).

High-growth rms are a small percentageof all new enterprises, but face distinctiveproblems (Anyadike-Danes, Bonner et al.2009).

Firms rarely experience continuous growth,discontinuous high-growth episodes are morecommon (Garnsey and Heffernan 2005a).

High-growth episodes usually result froma combination of internal rm factors andexternal factors in the business environment(Penrose 1959).

It is recognised that opportunities for high

growth are typically associated with uncertainty,as obvious opportunities for all leads to rapidexploitation which reduces the scope forentrepreneurial rents. As a result of uncertainty(e.g. technological, market, regulatory),selecting rms with potential for high growth isalso an uncertain process liable to errors:

...inventions can occur at any time,

with dierent importance and at varying

rhythms. Not all o them become

innovations and not all innovations diuse

widely. In act, the world o the technicallyeasible is always much greater than that

o the economically proftable, and this,

in turn, is much greater than that o the

socially acceptable. (Perez 2004) p.219

If high-growth rms do share thecharacteristics described above, then incubatorsare being asked to add value to an entity ina state of uncertainty. A portfolio approachmitigates the risk associated with relying ona singular rms performance in conditionsof heightened uncertainty.23 The portfolio oftenants in an incubator fall into three broadgroups: those for whom incubator support iscritical to improving rm performance, thosewhose development would likely occur withor without incubator support, and those forwhom the support is important (Figure 10).22Assessing when an incubatee needs critical helpas opposed to life support requires experienceby the incubator manager. Affecting outcomesin new venture performance depends on thetools of incubation available.

5.2.2 Delivery of valued servicesThe incubator building and facilities canthemselves be valued by entrepreneurs,

Figure 10: Additionality importance of incubator to company performance (modiedfrom CSES 2002). Arrows represent the ability of rms to move categories.

22. Venture investors alsotake a portfolio approachfor similar reasons. Fundmanagers accept thatlemons ripen before plumsand do not seek to prolongthe existence of thosecompanies that fail early;nor do they expend time ormoney on the living dead,those companies in theportfolio that neither thrivenor fail.

23. Anecdotal evidence suggests

incubator managers refer tothese three groups as: thosewho are unlikely to succeedregardless of help; thosewho are likely to succeedregardless of help; and thosefor whom help makes asignicant difference for achance of success.

Critical Important

60.6%22.5% 16.9%

Notimportant


23/54

3

24. For example, in a survey ofincubator services focusedon business assistance andnetworks, over a third ofrespondents attributedmajor value to rent breaks(43 per cent), businessconnections outside theincubator (41 per cent) andgovernment grants andloans (38 per cent) (Mian

1996). especially if designed for business incubation.The building address has been consideredan advantage to business for 79.6 per centof respondents in a UK survey (UKBI 2009).Incubatees preferred the incubator to be clearlyidentiable, which is aided by clear physicalboundaries around the incubator, for examplehaving a separate entrance (UKBI 2009). Thelook and feel of the incubation space can implythe quality of tenants and services it delivers.Resource providers can be wary of dealing withnew ventures lacking credibility and legitimacy(Bhid 2000). The incubator lends credibilityand legitimacy through association with theventure. Shared facilities enable incubatees touse professional facilities (e.g. meeting rooms,reception, ICT etc.) without the burden of beingwholly responsible for their cost. Incubationspace can also be designed to encourage peer-to-peer networking through the provision ofcommunal spaces, such as common rooms andcanteens, located in visible and accessible areas.Early studies of incubation emphasised facilitiesand administrative services, with more recentcontributions emphasising the importance

of business support and networks (Hansen,Chesbrough et al. 2000; Hackett and Dilts2004b).

The incubator delivers services and providesaccess to resources and networks typicallyvia its own incubator staff and externalconsultants. Typical incubator servicesand resources reect the needs of theentrepreneurial process (as shown in Figure9). For example strategic input to the businessmodel, access to resources including capital,organisational and recruitment support to buildthe productive and commercial base, access totechnical facilities, advice on capturing valuefrom innovation through intellectual propertyrights, and so on (Mian 1996; Hackett andDilts 2004b; Grimaldi and Grandi 2005; Patton,Warren et al. 2009).24 These services can bedelivered in varying degrees of quality, quantityand intensity. For example counselling is acritical function of many business incubators.A study of eight incubators in which incubatormanagers were questioned, in addition totenants nominated by the incubator managers,indicated that better performing incubatorshad proactive crisis intervention and proactivedevelopment intervention (Rice 2002). Whilean analysis of the type of counselling according

to whether it was episodic reactive, episodicproactive, or continuous proactive seemed tosupport the idea that proactive interventions

Figure 11: Link between counselling and incubator impact (amended)

Source: Rice 2002.

Hours per week dedicated to co-production Intensity (hours/intervention episode)

10

5

25

20

15

0

1

0.5

3

2.5

2

1.5

0

Hoursperweekdedicatedtoco

production

Intensity(hours/interventio

nepisode)

Incubator no. (increasing impact to right)

7 2 1 6 5 3 4 8


24/54

4

in incubatees generated more positive results(Figure 11), other studies suggest the bestincubatees (which were linked to high-growthepisodes) were also those least likely todemand help from the business incubator.

In addition to direct counselling and business

services delivered through the incubator, theincubator often acts as a mediator betweenthe entrepreneur and other resources andnetworks. It has been suggested that betterincubators offer an extensive network ofpowerful business connections that can betransformative to the development of itstenants (Hansen, Chesbrough et al. 2000). Ifan incubatee lacks its own access to relevantentrepreneurial networks this can be highlyvalued (Bergek and Norrman 2008). Networksplay a central role in the emergence andgrowth of successful rms (Hite and Hesterly

2001). When rms lack credibility the role of amediator to provide access to networks can beinvaluable. It is suggested that as rms becomemore established, their networks become morecalculated to t the increased quantity andscope of resource needs (Hite and Hesterly2001). The incubator can also offer institutionalmediation e.g. for access to public grants andprogrammes (Bergek and Norrman 2008).

It has been suggested that incubators varyalong three dimensions: selection strategies,

business support, mediation (which alsodepends on the regional the technologicalinnovation system) (Bergek and Norrman2008). Bergek and Norrman (2008) proposethat incubators fall into ve differentcategories where best practice will varybetween categories, though further work isneeded. A key nding was that even incubatorswith the same objectives can have differentdelivery mechanisms that are equally effective(Bergek and Norrman 2008). A similar attemptto categorise different incubators foundtwo broad types even though many of theincubators in the study exhibited featuresof both types (Grimaldi and Grandi 2005).Mian developed an integrative frameworkaccounting for various possible stakeholderssuch as a university, the entrepreneur,incubator management and community, but didnot weight their relative importance accordingto different modes and objectives of incubation(1997). So while the literature stronglysupports the existence of different approachesto incubation, no clear categories exist fromwhich to assess best practice within similar

cohorts of incubators.

5.3 Incubation can have a positive andcritical impact on new ventures

We have shown that business incubation canlead to a variety of outcomes in new ventures,and determining which ventures could havehigh-growth episodes is subject to uncertainty.

Nonetheless incubation can have a criticalaffect on improving the performance of newventures (Figure 10). While attempts havebeen made to discover best practice in businessincubation, it is also recognised that a varietyof incubation models exist. In short, organisingincubation to deliver maximum value fromtenants is contingent upon the internalsituation (including tenants) in the incubatorand external business environment which wenow explore in more detail.


25/54

5

The entrepreneurial process (Figure 9) isfundamental to technological change, and

yet is sometimes treated as a black box inthe innovation management literature (Jaffe,Newell et al. 2001). Yet it is this entrepreneurialprocess within the rm which builds acommercial structure around an innovationwith the aim of creating and capturing value(Stam and Garnsey 2005). The incubationprocess needs to support the entrepreneurialprocess through matching incubator services tothe changing needs of rms.

Demand for different types of incubatorservices varies according to thechanging needs of new ventures

6.1 New venture activity and businesssupport needs vary between regions

Regions can be associated with particularkinds of new ventures as a result of specicregional competences, resources and cultureas documented in literature on clusters. Manytechnology incubators are for example locatednear universities, and universities sponsoredsome of the rst incubators in the UnitedStates. Universities can boost the number ofnew rms appropriate for incubation via spin-outs25 and unofcial start-ups. A recent studyfrom the Cambridge cluster found unofcialstart-ups to outnumber spin-outs (Figure12) (Garnsey and Mohr 2010). In addition touniversities, research institutes and technologyintermediaries can also be valued sources ofstart-ups (Connell and Probert 2010).

Universities and other research institutionscan offer access to advanced technologylaboratories, equipment and other research

and technical resources, but also offer accessto talent such as faculty, staff and students

(Phillips 2002; Koh, Koh et al. 2005). In areview of incubator studies, universities havebeen identied as a key factor in the successof incubators (Hackett and Dilts 2004b; Phan,Siegel et al. 2005). However, it is suggestedthat the role of universities should be morethan just geographic proximity (Ratinho andHenriques 2010), and should include formaland informal exchange relationships with theincubator (Rothschild and Darr 2005). It isargued that people are the main contributionof universities to rms rather than specic

technologies, and that a people-centricapproach to the innovation process should beemphasised (Allott 2006; Connell and Probert2010).

A study of spin-out activity and entrepreneurialsupport strategies found three models(low selective model, supportive model,incubator model) largely dependent on theentrepreneurial orientation and supportstructures of the university and region.26 Asnot all universities have a cultural t withentrepreneurial endeavour, the low selectivemodel is involved in changing this cultureand often works with university researchersat the end of their contracts on very early-stage ideas requiring much support. Publicfunding is usually attracted due to the needfor enterprise to create jobs in areas with highunemployment. A more supportive model issimilarly positioned but with more interventionslike business plan competitions and a needfor business support in specialised unitse.g. incubators. The third incubator model,typical of regions like Cambridge, UK where

the university and region largely embraceentrepreneurial culture, are more focusedon the support of ventures complementary

Part 6: Matching incubator services to the changing

needs of rms

25. Academic spin-outs havebeen associated with lowerfailure rates if benettingfrom a strong IP licenseand/or have a professoron senior management,but graduation from theincubator is also retarded(Rothaermel and Thursby2005).

26. Clarysse, B. et al. (2005)

Spinning out new ventures:a typology of incubationstrategies from Europeanresearch institutions. Journalof Business Venturing. 20:183-216, p.194.


26/54

6

to regional expertise and with a globalcommercialisation perspective. In this thirdmodel, venture capital funding is oftenessential to enable ventures to fully develop.To adopt a model without consideration ofregional resources and competences risksfailure, as often happens when regions expectto emulate activities in atypical regions likeSilicon Valley or Cambridge, UK (Clarysse,Wright et al. 2005).

Similar results were found in incubator studiespre-2000. A signicantly lower technologicalorientation was found to exist in incubatees atrural incubators in Germany (Sternberg et al.1997 in Tamasy 2007). Luger and Goldstein(1991) propose that science parks in smallerregions (


27/54

7

levels and kinds of services dependingon their prior entrepreneurial experience(Figure 14). Companies without start-upexperience sought the most support withfunctional skills e.g. marketing, IT, legal and

government regulations in addition to marketand opportunity understanding. Companieswith start-up experience sought the mostsupport in strategic information e.g. marketand opportunities, customers, PR in additionto access to related R&D activity (Lacher2011). Further research is needed to assesshow generalisable these ndings are to otherincubators, and the implications of this forimproving incubator performance. Nonethelessit is likely that entrepreneurs benet frompeer-to-peer networking due to the mix inentrepreneurial experience between peers.

6.2 Matching incubator services to theneeds of tenants

As shown, the business support needs ofrms vary over time and across regions. Anincubators services should reect the needsof their clients i.e. rms. Evidence suggestsincubator services do vary and supportenterprise for a variety of goals e.g. specic

demographic groups (e.g. African-Americans,women), industries (e.g. fashion, food) orto stimulate enterprise in poor regions. The

dominant types of incubator programmesremain technology or mixed use, with 39 percent of US programmes being classied astechnology, 54 per cent as mixed-use (Knopp2007). Another study reported 45 per cent

of incubators as technology based with 36per cent being mixed use from a survey of78 incubators across 19 countries (MubarakAl-Mubarak and Busler 2010). A US studyfound technology programmes to offer moreintellectual property management compared toother programmes (Figure 15). More research isneeded to understand how much specialisationis appropriate for different business incubationcontexts. Typically a business incubator willintroduce some kind of selection mechanism topotential tenants if it is to specialise, which wenow discuss.

6.2.1 Selection processes of incubators

Rather than cater to all types of rms, mostbusiness incubators introduce a selectionprocess to target a particular group ofrms. There seems to be agreement amongresearchers that selection is an importantincubator management task (Hackett andDilts 2004b; Bergek and Norrman 2008). InEurope 97 per cent of incubators use a setof screening factors to evaluate potentialtenants (Aerts, Matthyssens et al. 2007).

Implementing appropriate selection processesfor entry into an incubator enables a bettert between the services it provides and the

Figure 13: Entry waves of rms in different sectors in Cambridge

Electronics

Manufacturing

Instrumentation

IT

Biotechnology

Telecommunication

R&D

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008


28/54

8

needs of tenants. It is a task subject to errorsowing to the challenge of distinguishingbetween the potential of entrepreneursoperating with different types of uncertainty(e.g. technological, market, regulatory etc.). Ithas been argued that this is reason enough tolet as many entrepreneurs try as is reasonablypossible since important qualities, for instanceif the person is coachable, are not possible to

fully detect at a screening meeting (Aaboen2009 p.661).

Selection processes can be broadly split intothose focused mainly on the idea or thosefocused primarily on the entrepreneur orteam (Bergek and Norrman 2008) (Box 2).Another study exploring the link betweenscreening practices and incubator performancesuggests a signicantly positive relationshipbetween tenant failure and the S-indexi.e. a high concentration on one screening

dimension (nancial factors, team or market)(Aerts, Matthyssens et al. 2007). This impliesscreening processes should include a variety

Figure 14: Comparison of services sought by companies with and without start-upexperience at St Johns Innovation Centre, Cambridge UK

Source: Lacher 2011.

94%18%

93%

93%

93%

89%

89%

89%

87%

86%

84%

83%

80%

80%

80%

80%

80%

80%

79%

79%

77%

74%

73%

71%

71%

71%

69%

69%

67%

67%

67%

64%

64%

63%

56%

56%

50%

53%31%

41%31%

36%17%

35%25%

33%27%

21%

56%

20%

29%

25%

33%

36%

11%

25%

17%

18%

33%

27%

30%

22%

20%

22%

40%

20%

44%

30%

20%

22%

31%

46%

27%

25%

64%

64%

57%

18%

25%

20%

Functional Skills: Marketing

Functional Skills: IT

Functional Skills: Legal/ Government Regulations

Coaching on Market and Opportunity Understanding

Strategic Information (in general)Strategic Information on Market and Oppportunities

Strategic Information on Customers

Functional Skills: Public Relations

Business Plan Assistance/ Strategy Formulation

Personal Recruiting/ Access to Labour Markets

Functional Skills (in general)

Capital/ Funding (in general)

Functional Skills: Finance

Functional Skills: Accounting

Functional Skills: Insurance

Coaching (in general)

Coaching on Growth Management

Technology Transfer Programmes

Library Services

University Related Resources/ Networks (in general)

Government Grants

Coaching on Leadership and Management Capability

Faculty Consultants

Coaching on Strategy

Coaching on Functional Skills

Labs and Workshops

Outside Capital:

Strategic Information on Suppliers

IP Assistance

Functional Skills: R&D

Related R&D Activity

Coaching on Innovation Management

Databases

University Finance

Distribution Channels

Angel Investors

Corporations

Manufacturing

Venture Capital

Bank Loans

Companies withoutstart-up experiencethat needed support

Companies withstart-up experiencethat needed support


29/54

9

Box 2: Selection strategies

Out of the incubators studied (16 Swedish VINNKUBATOR incubators) six were focused onthe entrepreneur, seven on the idea with three having equal emphasis on both. Most had anemphasis on picking the winners but also had pre-incubation processes with a qualicationprocess of ideas. Only one had a survival of the ttest approach where around 40 per centof candidates were accepted, though this incubator also had a signicantly higher numberof incubatees. Other incubators had a rejection rate of around 80 per cent. (ResearchMethodology involved screening incubator applications or the VINNKUBATOR programme

which has limitations). (Bergek and Norrman 2008)

Source: Bergek and Norrman 2008 p.24.

Selection strategies

Idea-focusedselection

Entrepreneur-focused selection

Survival of the ttest

The portfolio will presumably consistof a quite large number of ideaowners (or upcoming entrepreneurs)with immature ideas related to abroad spectrum of elds.

The resulting portfolio willbe diversied, and consist ofentrepreneurs/teams with strongdriving forces representing a broadset of ventures.

Picking the winners

Results in a highly niched portfolioof thoroughly screened ideas withina quite narrow technological area often sprung from the research ofhighly ranked universities.

The portfolio consists of a fewhandpicked and carefully evaluatedentrepreneurs, commonly withideas coupled to the research areasof a nearby university.

Figure 15: Percentage of incubators that offer these services across different programmes

Source: Knopp 2007.

Percentage

Linkages to strategicpartners

Technologycommercialisation

Intellectual property

management

Access to venturecapital investors

Access to angelinvestors or networks

Shadow advisoryboads or mentors

Management teamindentification

Linkages to highereducation resources

30 402010 10090807060500

Other programmes Mixed programmes Technology programmes


30/54

0

of factors. An alternative approach is takenby organisations like St Johns InnovationCentre that carefully position their brand toenable ventures to self-select if they thinkthe facilities and services are matched totheir needs. Like a venture capital investor,incubators need deal ow. A selection processcan only be imposed if the business incubatorcan afford to turn away potential tenants.Many incubators rely on rental income as partof their revenue stream and so need to llcapacity.

6.2.2 Graduation process

Graduating rms out of incubators createsroom for new incubatees. While someincubators retain anchor tenants, manyhave a policy for rm graduations. Howeversurveys indicate this is rarely strongly enforced(CSES 2002). It has been suggested that moreproactive monitoring of incubatees duringincubation may enable interventions to supportbusiness closures rather than prolonging timeto closure (Schwartz 2009). But rm closurestill remains a challenging decision to be made

by founders.

6.2.3 Continuity during uctuations inentrepreneurial support mechanisms

In addition to the changing needs ofentrepreneurs who request supportfrom incubators, incubators must alsoremain engaged with uctuations inentrepreneurial support mechanisms. Forexample, we have recently seen a switchfrom regional development agencies to acentralised Technology Strategy Board forthe administration of grants for new rmdevelopment in the UK. In addition to changesin public support mechanisms, the privatesector also varies over time as shown by thedramatic differences in the amount of equitynance available through venture capitalistsover the last three years (Figure 16). Theincubation environment offers an opportunityto reduce the search costs involved in stayingup to date with entrepreneurial supportmechanisms through access to incubation staffand peers. Incubators are able to help tenantstransition between different approaches andcapacities of varying entrepreneurial supportmechanisms.

Figure 16: UK Investment by nancing stage

Source: BVCA.

2009 2008 2007

3,000

4,000

2,000

1,000

8,000

7,000

6,000

5,000

0

Amount

invested

(m)

Seed

Startup

Earlystage

Laterstageventure

Expansioncapital

Bridgeloan

Replacementcapital

Refinancingbankdebt

PIPE

Secondarybuy-out

Other

Rescueturnaround

Publictoprivate

Managementbuy-out

Managementbuy-in


31/54

1

6.3 Without t there is failure

Tenants seem to become dissatised withincubator support when the incubatorsprogramme is predetermined rather than re-evaluated depending on the changing needsof tenants (Ratinho and Henriques 2010).27A prescriptive strategy enforced throughrigid monitoring can erode the ability toapply lessons learnt and adapt to changingconditions (Teece, Pisano et al. 1997). A criticalfunction of incubators seems to be the abilityto learn and adapt to the changing needs oftheir tenants:

It is important or us to stress the useulness

o a variety o dierent incubators adhering

to dierent incubating models, whose

rationale lies behind the existence o

companies with dierent business models

and with dierent requirements.(Grimaldiand Grandi 2005) p.119

Similarly it is important to take accountof regional differences when establishingincubators.28 For example while many try andemulate incubation strategies from Boston orSouthern California (US) or Cambridge (UK),these regions are also considered atypicaland likened to regional incubators owing tothe amount of entrepreneurial activity andinfrastructure (Clarysse, Wright et al. 2005).

In recent years, both funds and incubatorshave been used as policy tools for regionaldevelopment, especially the fostering ofinnovation or the commercialisation ofresearch. What was recently written of venturecapital may also be said of incubation: public

sector venture capital is unable to createentrepreneurial regions and [] a regionally-

based model o public sector venture capital

is ineective because it lacks scale. A new

approach or venture capital-defcient regions

is thereore required which gives greater

emphasis to the demand side [] The key

policy question is whether public sector venturecapital is an eective means o achieving

regional development. Emerging evidence is not

encouraging.(Mason and Perrakis 2009) pp1,24. Similarly placing an incubator in a regiondoes not guarantee it will have suitable tenantsto incubate nor attract sufcient support orresources from the local business environment.Neither venture funds nor business incubatorson their own can create an entrepreneurialor innovative ecosystem. To be successfulthey must work with a wide range of other

actors, from research institutions to (serial)entrepreneurs to specialist advisers, grant-providers, angel investors and many more.

Even within an entrepreneurial region it cantake time for incubators to become embeddedin the local business environment. The life cycleof an incubator starts when its establishmentis proposed (Allen 1988; Aaboen 2009).Once built, the incubator aims to achieve fulloccupancy and stable demand for space, nallyreaching a stage of more demand for spacethan it can service and becoming a centreof entrepreneurial gravity in the community(Allen 1988; Hackett and Dilts 2004b) (Figure17). Initially a young incubator is more likelyto suffer from insufcient demand for itsservices and fail to reach a critical mass of itstarget clients (Tamasy 2007). As an incubatorbecomes more embedded and known, therecruitment of new tenants should becomeeasier and with more potential tenants tochoose from the more selective the incubatorcan be (Aaboen 2009). As an incubatordevelops it should build knowledge and

networks that increasingly meet the needs oftenants in combination with the resources andopportunities associated with the local businessenvironment. This can lead to increasingspecialisation by the incubator.

Figure 17: Life cycle of incubator

Source: Adapted from Aaboen 2009.

27. Madeira Tecnopolosincubation programmeis a cause of tenantdissatisfaction. Companiesare admitted after winningan award that entitles

them to a years businessincubation services, butthe services are rigid andoften inappropriate for therms. In the same studythe Sogist Incubator wascriticised for locking tenantsinto a one year lease whichreverted to shareholdingcapital on the agreementof both parties. Successfulcompanies did not agreeto this while unsuccessfultenants needed an increasein capital. As a result Sogisthas not acquired any shares.(Ratinho and Henriques2010).

28. For example: Brazilianindustry associationsutilised the incubator toexpand clusters of low-techrms while municipalities,universities and industrialassociations combined theobjectives of the high- andlow-tech incubators ina mixed format. Publicuniversities and politicalgroups with social objectiveinvented the incubator forcooperatives as a means ofcombating poverty. P.412Etzkowitz, H., ManoelCarvalho de Mello, J. etal. (2005). Towards meta-innovation in Brazil: The

evolution of the incubatorand the emergence of atriple helix. Research Policy.34: 411-424.

No incubator

Identification of needs

and resources

Incubator established

Stakeholder

engagement

Commitment of

resources

Incubator embeds

Demand for

incubator stabilizes

Greater support for

incubator

Incubator matures

Demand for incubation

exceeds supply

Incubator a central part

of entrepreneurial

activity


32/54

2

While it has been possible to provide someindication of income and expense by incubators

and key business model variables, we ndfurther research is needed to understand thenancial implications of different models.

Experimentation with busines

incubation for growth

Documents