Content uploaded by Mike Wright
Author content
All content in this area was uploaded by Mike Wright on Jul 27, 2019
Content may be subject to copyright.
Understanding a new generation incubation model: The accelerator
Charlotte Pauwels
a
, Bart Clarysse
a,b,
n
, Mike Wright
a,b
, Jonas Van Hove
a,b
a
Universiteit Gent Faculteit Economie en Bedrijfskunde, Tweekerkenstraat 2, 9000 Gent, Belgium
b
Imperial College Business School Tanaka Building, South Kensington Campus, London SW7 2AZ, United Kingdom
article info
Article history:
Received 30 April 2014
Received in revised form
17 July 2015
Accepted 9 September 2015
Keywords:
Incubation models
Accelerators
Activity system perspective
Design
abstract
Prior research hints at the accelerator as a new generation incubation model. Accelerators have become
an umbrella term for any program providing a service structure of mentorship, networking opportunities
and access to funding. The challenge, however, is to understand their distinctive characteristics and
profiles geared towards reinforcing business start-ups. How do accelerators operate as a new generation
incubation model and how do they differ from existing incubation mechanisms? This inductive study
investigates 13 accelerators across Europe and adopts a design lens to identify the accelerator model’s
key design parameters. We identify five key building blocks and distinguish between three different
types of accelerators, taking the primary design theme of the accelerator into account. We contribute to
the incubation literature by extending recognition of the heterogeneity of incubation models, by deli-
neating the accelerator as a distinctive incubation model and by introducing the design lens as a useful
theoretical framework to investigate incubation models and their evolution.
&2015 Elsevier Ltd. All rights reserved.
1. Introduction
Over the past decades a wide variety of incubation mechanisms
have been introduced by policy makers, private investors, corpo-
rates, universities, research institutes etc. to support and accel-
erate the creation of successful entrepreneurial companies. Whilst
extant literature on incubation mechanisms agrees on their con-
tribution to the nurturing of new ventures in general, it also points
to the need to take the heterogeneity of different incubation
models into account (Barbero et al., 2014). Incubation models have
evolved (Bruneel et al., 2012) and continue to evolve into new
generation incubation models. It is therefore important to gain
insights into the specific features of evolving incubation models to
assess their working and performance (Mian, 1997) and their im-
pact on incubated ventures (Barbero et al., 2012).
A new generation incubation model, introduced in Europe in
the last five years, is that of the seed accelerator program. “Ac-
celerators”are organizations that aim to accelerate successful
venture creation by providing specific incubation services, fo-
cussed on education and mentoring, during an intensive program
of limited duration (Cohen and Hochberg, 2014;Miller and Bound,
2011). Accelerators emerged mid-2000 as a response to the
shortcomings of previous generation incubation models, which are
primarily focused on providing office space and in-house business
support services (Bruneel et al., 2012). The first accelerator, Y
Combinator, was established in 2005 in Cambridge, Massachusetts,
and has been a source of inspiration for many accelerators to fol-
low. In 2009, the Difference Engine kick-started the European ac-
celerator sector and in 2013, Seed-DB, a platform which analyses
accelerators and their companies worldwide, reported over 213
accelerators worldwide, which have supported approximately
3,800 new ventures.
Yet, despite these success examples and the rapid proliferation
of accelerators across different regions, empirical and theoretical
knowledge about the distinct characteristics and drivers of this
new generation incubation model is scant (Birdsall et al., 2013).
Furthermore, insights from the extant incubation literature only
partly help us to understand the working of accelerators. Research
on incubation models has provided in-depth insights into the
differences in the organization, activities, services and objectives
of incubator types (Aernoudt, 2004). However, we cannot simply
assume these differences hold for accelerators, which seem to
extend existing approaches to a very distinctive type of incubator.
In addition, the business incubation literature lacks a theoretical
lens to analyse and explain the heterogeneity among different
incubation models, with the majority of published studies being
largely descriptive in nature (Bruneel et al., 2012;Hackett and
Dilts, 2004).
Against this backdrop, we set out to explore 13 accelerators in
Europe in order to answer the following research question: “How
do accelerators operate as a new generation incubation model?”
Specifically, we introduce the design perspective developed by
Zott and Amit (2010) in their study about business models as a
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/technovation
Technovation
http://dx.doi.org/10.1016/j.technovation.2015.09.003
0166-4972/&2015 Elsevier Ltd. All rights reserved.
n
Corresponding author at: Imperial College Business School Tanaka Building,
South Kensington Campus, London SW7 2AZ, United Kingdom.
E-mail addresses: b.clarysse@imperial.ac.uk (B. Clarysse),
mike.wright@imperial.ac.uk (M. Wright).
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎
useful theoretical lens to look at the phenomenon and identify an
accelerator’s primary design parameters. This enables under-
standing of how accelerators differ from previous generation in-
cubation models and how they particularly create value for their
ventures. By doing so, we aim to contribute to the existing in-
cubation literature in two ways. First, by delineating accelerators
as a new generation incubation model. By identifying accelerators'
key design parameters, we conceptualize both the dimensions of
their heterogeneity and their distinctiveness in relation to other
incubation models. Second, by introducing a design lens as an
appropriate theoretical framework for investigating new incuba-
tion models, so enabling the consistent monitoring of incubation
model evolution.
2. Theoretical background
2.1. Incubation models
An incubation model is broadly defined as the way in which an
incubation entity provides support to start-ups to improve the
probability of survival of the portfolio companies and accelerate
their development. It is the model used by the organization or
mechanism to deliver incubation services to start-up companies
and create and capture value from them (Amit and Zott, 2001;
George and Bock, 2011). Incubation models have evolved since the
establishment of the first incubators, science parks, innovation
centres and the like. Academic research has followed this evolu-
tion by providing a variety of studies focusing on different in-
cubation model characteristics, classifications and typologies, and
their evolution over time.
2.1.1. Incubation model characteristics, classifications and typologies
The main body of research on incubation has devoted con-
siderable attention to describing different incubation mechanisms
and models (Barbero et al., 2014). The literature on academic en-
trepreneurship for example, focuses on how universities nurture
spin-offs into successful start-ups via internal approaches such as
technology transfer offices, science parks and incubation infra-
structures (Clarysse et al., 2005;Van Looy et al., 2003). The lit-
erature on corporate entrepreneurship illustrates how large com-
panies, similar to universities, rely on quasi-internal activities and
develop in-house incubation facilities to assist new start-ups as a
means to source new ideas (Becker and Gassmann, 2006;Grimaldi
and Grandi, 2005;Hill and Birkinshaw, 2014). In the public sector,
business incubators are recognized as a popular instrument to
foster entrepreneurship and regional economic development
(Smilor & Gill, 1986) and in the private sector incubation through
rent-seeking has grown into a separate industry, with the in-
volvement of investors as a way to improve the deal flow of their
portfolio (Miller and Bound, 2011). The latter is perceived as a
high-risk investment model for the support of high-potential new
ventures, originating from the venture capital and corporate
industry.
As incubation mechanisms have matured and multiplied, dif-
ferent incubation models have emerged, resulting in a plethora of
definitions and typologies, based on a variety of distinguishing
characteristics. The most fundamental categorization concerns the
distinction between non-profit and for-profit incubation models
(Aernoudt, 2004;Grimaldi and Grandi, 2005). Beyond this basic
dichotomy, research provided different classifications primarily
depending on strategic objectives, service offerings and competi-
tive focus, the latter distinguishing between industry sector, type
of start-up, phase of intervention and geographical reach (Van-
derstraeten and Matthyssens, 2012). Barbero et al. (2014) converge
on four broad models: (1) business innovation centres, with a
focus on regional economic development, (2) university incubators
to facilitate technology commercialisation, (3) research incubators
embedded in research institutes to valorise research output, and
(4) stand-alone incubators, focussed on selecting and supporting
high-potential ventures.
Previous research also identified a range of basic incubation
model components (Bergek and Norrman, 2008;Hackett and Dilts,
2004). Despite the differences and overlaps between incubation
models, an incubation model’s main components include at least
four of the five following services: (1) access to physical resources,
(2) office support services, (3) access to capital, (4) process sup-
port, and (5) networking services (Carayannis and von Zedtwitz,
2005), with a primary focus on overcoming the participating
venture's liability of newness, and hence improve its survival rate
(Dettwiler et al., 2006;Schwartz, 2013).
2.1.2. Incubation model evolution
A more recent stream of studies adopts a dynamic view on
incubation research, by focusing on the evolution of incubation
models over time (Grimaldi and Grandi, 2005). These studies ad-
vance the existence of a generational sequence of incubation
models, led by changing needs of participating ventures. They
argue that each generation of incubation models adapts its value
proposition to the evolving needs of participating ventures (Bru-
neel et al., 2012).
The first generation of incubation models, introduced in the
early nineties, primarily focused on providing physical and fi-
nancial resource support (for example office space and small fi-
nancial injections) to early-stage high potential ventures (Phan
et al., 2005). Throughout the nineties, new incubation models
emerged, which gradually moved away from a mere focus on of-
fering basic office space and financial support, towards a broad
range of more intangible high value added services. This second
generation of incubation models included, amongst other things,
services such as aid in evaluating different market opportunities,
access to knowledge intensive services, product development
support, access to knowledge, expertise and networks of en-
trepreneurs and provision of entrepreneurial finance (Clarysse and
Bruneel, 2007;Soetanto and Jack, 2013). More recently, we can
identify a further shift, hinting at a new generation of incubation
models, which focuses on knowledge intensive business services,
moving away almost entirely from the primary services for which
the incubation models were founded (i.e. rental services).
2.1.3. The accelerator: a new generation incubation model?
The accelerator model is an exemplar of the recent shift to-
wards a focus on intangible, knowledge intensive, support services
in incubation services. An accelerator is an organization, which
aims to accelerate new venture creation by providing education
and mentoring to cohorts of ventures during a limited time (Cohen
and Hochberg, 2014). Although the accelerator model includes
intangible services, such as mentoring and networking, it has a
number of other specific features that sets it apart from existing
incubation models (Isabelle, 2013). First, they are not primarily
designed to provide physical resources or office support services
over a long period of time. Second, they typically offer pre-seed
investment, usually in exchange for equity. Third, they are less
focused on venture capitalists as a next step of finance, but are
more closely connected to business angels and small-scale in-
dividual investors. One of the reasons for this difference is that
their focus is on early-stage tech start-ups for which the costs of
experimentation have dropped significantly in the last decade,
rather than capital-intensive start-ups, such as technology-or-
iented spin-offs from universities. Fourth, the accelerator model
places emphasis on business development and aims to develop
start-ups into investment ready businesses by offering intensive
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎2
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
mentoring sessions and networking opportunities, alongside a
supportive peer-to-peer environment and entrepreneurial culture
(Christiansen, 2009). Fifth, the accelerator model concerns time-
limited support (on average 3–6 months), focused on intense in-
teraction, monitoring and education to enable rapid progress, al-
though some provide continued networking support beyond the
program as well.
Although literature suggests that the accelerator model can be
considered a new generation incubation model (Wise and Valliere,
2014), formal analysis about its particular characteristics and dri-
vers is lacking. The few available studies examining accelerators
are largely descriptive in nature and lack a consistent theoretical
lens to study the phenomenon (Cohen and Hochberg, 2014;Miller
and Bound, 2011). We address this gap by providing a more in-
formed image of new generation incubation models in general and
the accelerator model in particular, as part of a broader effort to
introduce the design lens as a systematic methodological approach
to study incubation evolution.
2.2. A design lens to study incubation model evolution
The design lens introduced by Zott and Amit in their research
about business model design (Amit and Zott, 2012;Zott and Amit,
2007,2010) is a useful framework to study incubation model
evolution. This stream of research introduces the concept of an
organization’s activity system, concerning the set of inter-
dependent organizational activities conducted by the focal orga-
nization and its partners, enabling the organization to create, de-
liver and capture value in concert with these partners. It suggests
two sets of design parameters that should be taken into account
when choosing the appropriate “model”or “template”for the
activity system to perform its activities: design elements and de-
sign themes. Design elements are the key building blocks of the
activity system's model, which set it apart from other models.
Design themes represent the common theme that orchestrates
and connects the different elements into a particular model and as
such categorize different models of activity systems (Amit and
Zott, 2012).
The activity system design perspective is particularly relevant
to study a new generation incubation model, as it provides a
conceptual toolbox to identify and assess its key elements and
themes. It can be used to, on one hand, distinguish the new model
from existing models, through identifying the model’s vital ele-
ments and, on the other hand, reveal the heterogeneity within the
new model, through identifying the main themes characterizing
different types within the new generation model. As such it pro-
vides a structured framework for incubation researchers to con-
sistently track and assess incubation model evolution.
3. Methods: sample, data collection, and analysis
Given the lack of previous research specifically on accelerators,
the contemporary and therefore still relatively unexplored subject
under study, and our “how”-research question, we choose an in-
ductive, multiple case study design as a research strategy (Ei-
senhardt and Graebner, 2007;Tracy, 2010).
3.1. Sample
We use a theoretical snowball sampling approach (Yin, 2013).
This means that we started our sampling by only focusing on cases
that comply with a predefined strict definition of an accelerator,
clearly delineating the accelerator model from other incubation
models. Based on Miller and Bound (2011) we define an accel-
erator as having the following six characteristics: (1) Possible offer
of upfront investment (d10k–d50k), often in exchange for equity
(5–10%); (2) Time-limited support, comprising programmed
events and intensive mentoring; (3) An application process that is
“in principle”open to all, yet highly competitive; (4) Cohorts or
classes of start-ups rather than individual companies; (5) Mostly a
focus on small teams, not individual founders; (6) Periodic gra-
duation with a Demo Day/Investor Day. Using the above criteria,
we identified an initial dataset of 41 accelerators in Europe that
complied with our strict accelerator definition. We further im-
posed two additional criteria on the dataset to result in a final
selection of 14 cases (a) the cases are viewed by experts who sit in
the European accelerator advisory board, called the Accelerator
Assembly, as accelerators which have developed a track record and
have signalled to stay in the field for a longer time period and
(b) they are located in one of the three “leading accelerator re-
gions”in Europe: London, Paris and Berlin. The Regional En-
trepreneurship and Development Index (REDI), a complex com-
posite indicator of regional entrepreneurship that captures both
individual-level actions as well as contextual influences such as
the financial possibilities of businesses, ranked the regional en-
trepreneurial performance of London, Ile-de-France and Berlin
amongst the top in the European Union (Szerb et al., 2014). These
three cities created the conditions for accelerators to take off as
they have a sufficiently dense population of entrepreneurial ven-
tures to be attractive for accelerators and have a developed seed
stage funding supply resulting in better circumstances for start-
ups and start-up programs to make an impact (Salido et al., 2013).
We argue that focusing on the best performing accelerators only
contributes to our theoretical sampling approach as it facilitates
access to rich insights about an accelerator's key design para-
meters. As the accelerator model is still very young (average age of
3 years) we relied on expert judgements rather than established
performance indicators in incubation research such as the number
of jobs created, number of graduates and occupancy rate (Barbero
et al., 2012). Among the 14 selected, the managing directors of 13
accelerators agreed to participate in our study. Table 1 provides a
final list of the 13 accelerators included in the study and their key
characteristics.
3.2. Data collection
We used two data sources: interviews and archival data. The
primary data source involved semi-structured interviews with the
managing directors of the 13 accelerators selected, during the
second half of 2013 and early 2014, using the repertory grid
method as a technique to structure the interviews (Easterby-Smith
et al., 1996). The repertory grid technique focuses on the con-
struction of meaning by individual participants in a specific setting
and was chosen as a technique to supplement standard interview
questions, (such as “Can you describe your ideal portfolio com-
pany?”“
What makes your accelerator unique? etc.), due to its
comparative efficiency and flexibility and its greater potential for
objective validity and reproducibility (Symon and Cassell, 1998).
Interviews ranged from 50 minutes to 1.5 h and always involved
two researchers: one conducting the interview, and the other
taking field notes. Each interview was tape-recorded and tran-
scribed, which resulted in 215 pages of total interview transcripts.
The French-speaking interviewees were interviewed in their mo-
ther tongue, transcribed in French and then translated into
English.
The interview data was supplemented with archival data from
various sources, including industry reports, internal accelerator
program records, company presentations, annual reports, websites
and news articles about the organisation. These secondary data
sources were important sources of information to familiarize with
the context and construct preliminary case histories of each
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎ 3
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
accelerator, as well as served as triangulation sources to validate
emerging insights from the interviews (Huberman and Miles,
1983).
3.3. Data analysis
Our data analysis evolved in three stages. We started with
writing individual case histories of each case using all archival data
available. We then contacted the managing directors of the ac-
celerators through email to ask for an interview, with the pre-
liminary case history of their accelerator attached, in order to in-
crease response rates (Yin, 2013). Further communication through
email and telephone was used to schedule interviews and validate
the preliminary case histories.
Once the case histories were validated and interviews were
scheduled, we proceeded with conducting the interviews, using
the repertory grid method both as a data collection and data
analysis technique. We followed the three stages of the basic re-
pertory grid technique (Easterby-Smith et al., 1996). First, we de-
fined 15 accelerators (the 13 cases under study together with the
2 pioneering accelerators in US: Y Combinator and Techstars US) as
our “grid elements”(¼objects of attention within the domain of
investigation). Each grid element was written down onto an in-
dividual card. Second, we used “triads”and “the full context form”
(Tan and Hunter, 2002) as two techniques to elicit “constructs”
(¼qualities describing and differentiating elements). During the
first part of the interview we constructed a triad by combining the
interviewee’s own accelerator with two accelerators, randomly
drawn from the pack of cards. The three cards were presented to
the interviewee, who was then asked to identify ways in which
two accelerators are similar yet different or opposite from the
third. This process was repeated until no new constructs could be
identified. In the second part of the interview, we presented the
full repertory of cards to the interviewee and requested him or her
to sort the stack of cards into any number of discrete piles based
on whatever similarity criteria the interviewee chose to apply.
After the sorting was completed, the interviewee was asked to
provide a descriptive title for each pile of elements. Finally, after
completion of each interview, we constructed a “grid”(¼matrix)
of grid elements and constructs and completed each cell of the
grid with information from the interview (i.e. for each accelerator
we entered data in the cells representing how the accelerator is
regarded in terms of the identified constructs).
The third stage of our data analysis involved a cross-case ana-
lysis. As suggested by Eisenhardt (1989), one tactic in cross-case
analysis is to select categories and dimensions and then to look for
inter-case similarities and differences. The categories and
dimensions were suggested by the elements and constructs from
the grids built up for each interview and all cases were replicated
against one another (Yin, 2013). We counted an initial number of
17 constructs identified by the interviewees and applied two
rounds of comparative analysis to cluster constructs “that go to-
gether”(Miles and Huberman, 1994).
Afirst round of analysis resulted in grouping the 17 constructs
together in 9 elements. After a second round we eventually agreed
upon a final set of 5 design elements. The final set of 5 design
elements were reviewed by the interview respondents to further
validate our results. We finalized our analysis by identifying
themes cutting across cases. The full context form technique ap-
plied during the interviews resulted in a number of different
groups of accelerators, ranging from 2 to 5 different groups. We
further compared all of the data available for each case in a matrix
to reveal element relationships and agreed upon three distinct
groups of accelerators in our dataset. The three different groups
were again reviewed by the interview respondents to validate our
findings. We employed an insider–outsider approach, which
means that a third person was involved in the analysis rounds as
an independent researcher so that the credibility of the findings
would not rely solely on the interpretations of those conducting
the interviews (Gioia et al., 2010).
4. Findings
This section reports the results from the repertory grid con-
struction and cross-case analysis. We discuss the five accelerator
design elements and three accelerator design themes that emerged
from our findings.
4.1. Design elements
The design elements of an activity system capture the key
parameters that describe the activity system's architecture (Zott
and Amit, 2010). As outlined above, the 5 design elements of the
accelerator model were identified through comparative analysis of
the 13 cases involved, which led us to cluster the 17 constructs
identified by the interview respondents into 5 agreed upon design
elements. Fig. 1 and Table 2 illustrate how we arrived at the final
selection by respectively showing which of the 17 constructs were
clustered together in a design element, and portraying supportive
quotes for each of the 17 constructs. In what follows we describe
each design element in detail.
Table 1
Case descriptives.
Name Acronym Location Founding date Program length Investment size Equity stake
taken
1 Techstars London TL UK, London 2013 3 months d12500þoption conv. loan 6%
2 Healthbox Europe HB UK, London 2012 4 months d50000 10%
3 Fintech Innovation Lab FIL UK, London 2012 3 months / /
4 Bethnal Green Ventures BGV UK, London 2011 3 months d15000 6%
5 Climate-KIC Europe CKE Europe 2010 12–18 months Max. of €95000 /
6 Microsoft Ventures Accelerator MVA Germany, Berlin 2013 4 months / /
7 Axel Springer Plug and Play
Accelerator
ASPP Germany, Berlin 2013 3 months €25000 5%
8 ProSiebenSat.1 Accelerator PSSA Germany, Berlin and
Munich
2013 3 months €25000 5%
9 Startupbootcamp Berlin SBC Germany, Berlin 2012 3 months €150 00 8%
10 Le Camping LC France, Paris 2010 6 months €4500 /
11 The Family TF France, Paris 2013 indefinite / 3%
12 L’Accélérateur LA France, Paris 2012 4 months €10000þoption for more 7–10%
13 Scientipôle Initiative SI France, Paris 2002 6 months €2000090000 /
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎4
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
4.1.1. Program package
The program package consists of all services the accelerator
offers to its portfolio ventures. The accelerator program package’s
core services that most differentiate the accelerator from previous
generation incubation models are the well-elaborated and care-
fully planned mentoring services. Mentors are typically experi-
enced entrepreneurs, which are heavily vetted before being in-
cluded in the accelerator program. They are matched to specific
ventures based upon speed dating or match making events and are
frequently evaluated by the accelerator management team. Men-
tors help ventures to define their business model and to connect
with customers and investors. Although there are variations in
how this mentoring is operationalized, mentoring services are
evident across all accelerators.
An accelerator’s program package most often also includes a
curriculum or training program, covering a variety of topics such
as finance, marketing and management, which the new ventures
have to go through when entering the accelerator program. The
ProSiebenSat.1 accelerator for instance includes courses in finance,
user design, PR, marketing and legal aspects, and a program of ad
hoc events, such as, expert workshops and inspiring lectures.
In addition to educational services, accelerators offer regular
counselling services, provided by the accelerator management
team. These are offered in the form of weekly “office hours”or
evaluation moments and provide the portfolio companies with
business assistance and enable monitoring of their progress.
The portfolio companies are also given the opportunity to come
into contact with customers and investors through the organiza-
tion of demo days or investor days. During these days, customers
and/or investors are invited to visit the accelerator and attend
portfolio companies’presentations, followed by formal and in-
formal networking opportunities.
Location services are also part of the accelerator program
package, but are limited to co-location in a shared open office
space, with the aim to encourage collaboration and peer-to-peer
learning.
Finally, the program package also consists of investment op-
portunities offered to the portfolio companies. We find that most
programs (8 out of 13) follow the traditional accelerator model of
offering a small amount of funding in exchange for equity (ranging
from d3,600–d50,000 for 3–10%). The equity stakes are typically
made on a dilutable basis with pro-rata investments in ensuing
rounds being optional case-by-case. Some form of follow-on
funding can be provided as well. For example, Healthbox Europe
has shaped an Angel Fund that acts as a co-investment fund to be
invested alongside the accelerator as a separate legal entity.
4.1.2. Strategic focus
The second design element of an accelerator is the strategic
focus. The strategic focus concerns the accelerator’s strategic
choices regarding industry, sector and geographical focus. The
industry and sector focus ranges from being very generic (no
vertical focus at all) to very specific (specialized in a specific in-
dustry, sector or technology domain). For example, Fintech In-
novation Lab focuses exclusively on the financial sector, while
L’Accélérateur is more broadly “retail-oriented”. Overall, accel-
erators seem to be focusing their programs increasingly on certain
themes rather than being generic.
In addition to an industry and sector focus, accelerators also
have a geographical focus where they choose between being lo-
cally versus internationally active in their activities. Techstars is an
example of a program that initially focused on US only, but then
internationalized to Europe with a program in London and Berlin.
However, each local program operates autonomously, while
Techstars as a whole aims to share best practice across its local
units.
4.1.3. Selection process
Accelerators make use of a rigorous, multi-staged selection
process. Usually, an open call is organized for a period of time,
during which portfolio companies can register and apply online on
a software platform such as F6S.com, Fundacity or Angel.co. Some
programs, like Startupbootcamp and Climate-KIC, go one step
further and actively scout start-ups during events before the ap-
plication period.
Then, a standardized screening process is organized in which
external stakeholders tend to participate. Different types of sta-
keholders are asked to sit in a selection committee or to do in-
terviews. The portfolio companies are expected to present their
ideas and they are screened in person. For example, Healthbox
Europe uses a selection committee, which comprises of mentors,
investors and alumni, to help shortlist companies in its program.
Program
package
Mentoring
services
Curriculum /
training program
Counseling
services
Demo days /
Investor days
Location
services
Investment
opportunities
Strategic
focus
Industry / sector
focus
Geographical
focus
Selection
process
Online open call
Use of externals
for screening
Team as primary
selection
criterion
Funding
structure
Investor funding
Corporate
funding
Public funding
Alternative
revenues
Alumni
relations
Alumni network
Post program
support
Fig. 1. Design elements and constructs.
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎ 5
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
Table 2
Data structure supporting accelerator design elements.
Design elements Constructs Representative quotes
PROGRAM PACKAGE Mentoring services “80–100 individuals in our mentor network”[HB, Nov 2013]
“The mentor model came from Techstars US. Mentors are heavily involved in the program.”[BGV, Oct 2013]
“Start-ups are given feedback all the time, there is a structured feedback process regarding partners and
mentors.”[MVA, Dec 2013)
“We meet every mentor face-to-face and kind of have a debrief or pre-brief.”[SBC, Dec 2013]
“The only method that we found that works is: rent a room in a restaurant, bring in food and a lot of alcohol
and close the doors, and in 4 h magic happens”[MVA, Dec 2013]
“First month is mentor-heavy, with matchmaking and presenting and speed dating…”[SBC, Dec 2013]
“We have intern mentors, so from within the company, that have expertise in a certain area.”[PSSA, Nov 2013]
“We have godfathers…They are actually internal coaches from Axel Springer. So we match them with the
teams.”[ASPP, Dec 2013]
Curriculum / training program “We have like lawyers, accountants, and HR people that also offer their services to our start-ups through
workshops, lectures or office hours. Then we have some lectures that inspire them.”[MVA, Dec 2013]
“…fixed curriculum points they have to, or should attend. And those are sessions with internal and external
mentors and coaches, with experts, with entrepreneurs, with people from the team where they learn things
about specific functional topics.”[PSSA, Nov 2013]
Counselling services “We check with the companies at least weekly if not twice a week so we do have regular conversations.”[HB,
Nov 2013]
“We also do a kind of weekly stand-up. On Friday they have to stand in front of the class explaining what they
did last week and what they want to achieve.”[MVA, Dec 2013]
“We set up an action plan and use the cash to address the bottlenecks. To identify the bottlenecks, you need to
sit around the table for hours, maybe days…Then we have to follow-up by visiting the guy (founder) step-
by-step.”[KIC, Dec 2013]
Demo days / Investor days “Our Demo Day is slightly different. It is not about getting investors in the room, it is actually getting customers
in the room for the companies.”[HB, Nov 2013]
“The majority of people we invite for the Investor Day are investors and they could be angel investors, VC’s,
private equity investors…”[FIL, Nov 2013]
Location services “Free office space here, free Wi-Fi, free stunning view, free drinks.”[MVA, Dec 2013]
“We ask them to come to London and we provide them with desk space and office space.”[FIL, Nov 2013]
Investment opportunities “The deal is 100% standardized because we don't have time to negotiate with the teams…so we take 5% of
equity in the companies and we give them €25,000 plus our mentoring, coaching and the office space.”[PSSA,
Nov 2013]
“We invest some cash in the beginning. Between 5 and 15K. But if we believe that the companies are in the
right track and need some money then we will invest between500 and 200 K and we usually take between 7–
12%.”[LA, Nov 2013]
“So we have $120,000, $20, 000 dollars goes for 6% plus the program, plus all the freebies which are not
insignificant. And alongside that the teams get $100,000 on a note, convertible note.”[TL, Jan 2014]
“After graduation, we have the discretion of writing the 150,000 check. The alternative, which we do use a lot,
is we basically say we will co-match.”[TL, Jan 2014]
“We can do follow-up investments…if anyone comes and says like ‘I like them’and he invests, we can give the
other 50%. So we can mirror the investment.”[ASPP, Dec 2013]
STRATEGIC FOCUS Industry / sector focus “Nesta’s investment themes…health, education and sustainability”[BGV, Oct 2013]
“We are an open thematic accelerator”[LC, Jan 2014]
“We customized that model to be more reflective of the healthcare market and our interest in exploiting that
area.”[HB, Nov 2013]
Geographical focus “90% of our businesses are in the US and 10% is not…I used to run a stand-alone program and it would have
been hard to differentiate myself.”[LC, Jan 2014]
“The London program is very much based on the New York program that we have”[FIL, Nov 2013]
“There are these that are considered innovative enough since we only fund innovative projects that are less
than 3 years old and are in the region of France.”[SI, Dec 2013]
SELECTION PROCESS Online open call “We have an application phase that is open for about 4–6 weeks. During these 4–6 weeks we ask companies to
submit and to fill in a questionnaire”[PSSA, Nov 2013]
“We open the online platform for two months. So future applicants have two months to register and to
complete their applications”[LC, Dec 2013]
Use of externals for screening “We shortlist companies with the help of members of the selection committee who are representative of the
mentors and some of the investors of the program”[HB, November 2013]
“We use alumni a lot when screening, especially if the idea is in line with their area of expertise…I even let
them do interviews. And we do have a selection committee –theyare involved in the selection days”[SBC, Dec
2013]
“The banks meet the start-ups and they interview them”[FIL, Nov 2013]
Team as primary selection
criterion
“We learned that selecting teams remotely is difficult, we want to see them face-to-face, in action”[SBC, Dec
2013]
“We do a final panel interview which we do in person rather than Skype, because we want to meet the team”
[TL, Jan 2014]
“We have a focus when we look at selection: team, team, team and opportunity”[TL, Jan 2014]
“We have like 3 important criteria: the team, degree of innovation and market opportunity”[LC, Nov 2013]
“We look at personal qualities (ambition, tenacity, frugality, openness, flexibility) and strong teams which
interact well”[LA, Nov 2013]
FUNDING STRUCTURE Investor funding “We are privately funded mostly by business angels and a couple of VC’s”[SBC, Dec 2013]
“Our investors are either all professional investors or VCs orangels. And we cap the amount of money that any
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎6
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
Remarkably, all accelerators in our sample claimed that teams
are the main selection factor. Entrepreneurial teams are typically
selected in batches and single founders are only selected by ex-
ception. A representative example is the screening process of the
Paris-based accelerator TheFamily. Their application process is
perceived to be “founder-friendly”, since the team as opposed to
the idea is the dominant decision factor for participation in the
accelerator. Some accelerators will help founders with match-
making and team formation, which is also of benefit to teams
missing a specific skill set. For example the Paris-based accelerator
Le Camping organizes an event called “Adopt a CTO”before
opening the call to submit applications. This event offers single
founders the opportunity to find a CTO and form a team. Other
accelerator programs such as Startupbootcamp and Climate-KIC
have entrepreneurs-in-residence. These are entrepreneurs with a
specific skill who can join entrepreneurial teams, become co-
founders, or build their own companies. They are more than mere
advisors (compared to mentors), as they work closely together
with the teams and become team members. Some entrepreneurs-
in-residence are paid, others participate in the program driven by
the opportunity, experience or personal growth.
4.1.4. Funding structure
A fourth design element characterizing an accelerator concerns
its funding structure. We find that most programs receive the
major part of their working capital from shareholders. These
shareholders are either private investors, corporate companies or
public authorities. Although most accelerators look to complement
these sources with revenues, few of the accelerators we inter-
viewed were able to get revenue from investments in the start-ups
they support. Alternatively, this can also be because these pro-
grams are still relatively new and it will take some time before
they have noticeable exits in their portfolio companies. Some ac-
celerators diversify their model in order to source alternative
revenue through the organization of events and workshops. For
example, TheFamily organizes a lot of events, for which they sell
tickets online, which has turned into a profitable event business.
4.1.5. Alumni relations
The last design element particular for an accelerator concerns
its relations with alumni. The accelerators in the study put a lot of
emphasis on keeping close and active relations with the compa-
nies that graduate from their program. Most accelerators run
regular events for alumni and invite them back into the program
to share their experiences where possible. These companies are
used as reference cases and often get actively involved in the
mentoring activities discussed above.
Some accelerators experiment with the extended provision of
support services to alumni companies once graduated. Accel-
erators that take equity in their start-ups have an additional in-
centive for providing continued support to help their alumni
succeed. Once an accelerator has developed over a number of
years, the alumni network can be an important source for mentors
and investors, as successful graduates are more likely to invest
back into the community that supported them in the first place.
We conclude from our analysis that the five design elements –
program package, strategic focus, selection process, funding
structure and alumni relations –are the key building blocks of an
accelerator model, as they appear in each of the 13 cases under
study and allow parallels to be drawn and differences among the
cases to be identified.
4.2. Design themes
Our data further reveals that the accelerators in our study vary
in their architecture, depending on their approach to each of the
design elements. In the next section we therefore describe the
second set of design parameters that characterize an accelerator:
its design theme. The accelerator’s design theme is the common
theme underlying a particular type of accelerator, orchestrating
Table 2 (continued )
Design elements Constructs Representative quotes
investor can put into our fund. Because we actually want diversity in our investor base rather than 1 person
turning up and say ‘here is half the money’. So I tend to use it much more aggressively than some others do to
create a network of smart investors”[TS, Jan 2014]
Corporate funding “Accenture covers the operating costs”[FIL, Nov 2013]
“Then you have the ones that are corporate funded (like us), which is typically a prerequisite for providing a
good program that will last for a longer period of time”[PSSA, Nov 2013]
“There is no partner funding, so this is all Microsoft funded. There is no partnership with any organisation. I am
a 100% Microsoft employee, this building is financed by Microsoft.”[MVA, Dec 2013].
Public funding “It is a non-profit association and it is a sponsorship. So we receive some money and we allocate it, this money,
to our events and our place”[LC, Nov 2013]
“Wayra UnLtd is, like us, funded from the Cabinet Office…We have a non-profit part which owns the majority
of Bethnal Green Ventures LLP”[BGV, Oct 2013]
Alternative revenues “Actually we have a very profitable event business. We are organising a lot of events and people like our events.
So we know how to sell tickets online, it is a good way to gain money, the event business is an incredible
business with capital”[TF, Nov 2013]
“Startupbootcamp Berlin is renting out desks in our new co-working space called the Start-up Gallery”[SBC,
Dec 2013]
ALUMNI RELATIONS Alumni network “We build the infrastructure to try to help them…We run alumni-events quite often and alumni are invited
back in for all the program stuff when we run a program. So we create a lot of opportunities for them”[BGV,
Oct 2013]
“We have an alumni annual meeting where we bring as manyalumni as possible together. And they just share
what is going on and they connect from across the programs”[SBC, Dec 2013]
Post program support “Our program runs from October to January but we continue to offer office space until past September. So it is
one less thing for the companies to worry about because, you know, office space in London is extremely
expensive. So we continue to make introductions and continue to support the companies where we can.
Obviously it is not as hand-on as it was during the program but there is additional support”[HB, Nov 2013]
“We don’t kick the alumni out of our space, why would we? And we run monthly alumni events in London.
There is one tomorrow, every first Thursday of every month. We have it in the same space all the time”[TL, Jan
2014]
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎ 7
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
Table 3
Data structure supporting accelerator design themes.
ECOSYSTEM BUILDER DEAL-FLOW MAKER WELFARE STIMULATOR
Design theme “Matching customers with start-ups and build corporate
ecosystem”
“Identification of investment opportunities for investors”“Stimulation of start-up activity and economic
development”
Program package Mentoring provided by internal coaches from corporates No
seed investment or equity engagement
Mentoring provided by serial entrepreneurs and business angels-
Standard seed investment and equity engagement
Mentoring provided by serial entrepreneurs and business de-
velopers; most extensive curriculum Mostly seed investment
and equity engagement
Strategic focus Mix of generalists and specialists International focus Mix of generalists and specialists Local and/or international focus Mostly generalists Local and/or international focus
Selection process Favour new ventures in later stages with some proven track
record
Favour new ventures in later stages with some proven track record Favour very-early stage new ventures
Funding structure Funding from corporates Funding from private investors (business angels, venture capital funds
and/or corporate venture capital)
Funding from local, national and international schemes; ex-
perimenting with funding structure and revenue model
(search for sustainability)
Alumni relations Establish infrastructures to build alumni services Establish infrastructures to build alumni services Establish infrastructures to build alumni services
Cases Fintech Innovation Lab Microsoft Ventures Accelerator Techstars London TheFamily Startupbootcamp Berlin ProSiebenSat.1
Accelerator Axel Springer Plug and Play Accelerator L’Accélérateur
Climate-KIC Europe Scientipôle Initiative Le Camping
Healthbox Europe Bethnal Green Ventures
Representative quotes “It is more a service to strengthen our relationships with the
banks”[FIL, Nov 2013]
“The goal is to generate positive returns from our investments”[PSSA, Nov
2013]“We do it because we really would like to have a good investment
case…So when I look back in 8 years, I would like to have two big exits
because then everything we did here is fine. We help them with contracts,
follow-up investment, so we are also investment bankers.”[ASPP, Nov 2013]
“Get the economy going with social impact start-ups. It’s not just
about investing in start-ups”[BGV, Oct 2013]
”With Microsoft you have unparalleled access to customers, be-
cause we are still relevant and big in every small and midsize
enterprise.”[MVA, Nov 2013]
“The most important thing is to create sustainable start-ups in
the long term…about 200 jobs have been created”[LC, Dec 2013]
“We want to create more exit opportunities…we are privately funded by
investors”[SBC, Nov 2013]
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎8
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
and connecting the different design elements (Zott and Amit,
2010). The accelerator design themes were identified through
application of the “full context form”repertory grid technique
during the interviews (see above) and a further cross-case analysis,
focused on revealing themes cutting across cases. As explained in
the data analysis, our data revealed three distinct themes char-
acterizing three different types of accelerator. Table 3 provides an
overview of the different accelerator types, outlining the differ-
ences and similarities regarding the 5 design elements, and illus-
trates which cases belong to each group. In what follows we again
describe each type of accelerator in detail.
4.2.1. The “ecosystem builder”
The “ecosystem builder”is an accelerator typically set up by
corporate companies that wish to develop an ecosystem of cus-
tomers and stakeholders around their company. Large companies
such as Microsoft and Accenture install or support an ecosystem
builder accelerator in order to extend their network of stake-
holders. The accelerator is used as a matchmaking device to con-
nect lead customers with promising start-ups and in this way
nurture the development of an ecosystem around the company. As
an example, the accelerator FinTech Innovation Lab in London is
run by Accenture. It has the primary aim to create a platform for
the financial services industry to collaborate on innovation with
early-stage ventures. With this, Accenture seeks to strengthen its
relationship with banking clients and increase its foothold in the
market. Similarly, one of the drivers of the Microsoft Ventures
Accelerator is to support start-ups whose solutions will benefit
Microsoft's vast SME customer base across Europe.
The ecosystem builder accelerator actively involves its corpo-
rate stakeholders in the accelerator’s operations. For example,
senior executives of the corporate are often involved in the se-
lection process of portfolio companies. Hence, only those ventures
that attract the attention of the corporate’s executives and that
will be able to enhance the corporate’s ecosystem development
are selected. Mentors are often sourced from the corporates as
well. These corporate mentors help the start-ups to find their way
through the internal decision-making system of the company. In-
terestingly, this type of accelerator most often has no profitor-
ientation and offers no investment to the start-ups that participate
in the program. Instead, these accelerators add value to the port-
folio companies, primarily by helping them to connect with po-
tential customers. The accelerator’s network is therefore almost
exclusively oriented towards the potential customer base. They are
financed on a yearly basis by the corporate and often adopt soft
performance measures. They frequently engage in symbolic ac-
tions such as broadcasting, newsletters, and showcase events, to
illustrate their legitimacy in the absence of strict key performance
indicators (Zott and Huy, 2007).
4.2.2. The “deal-flow maker”
The “deal-flow maker”accelerator receives funding from in-
vestors such as business angels, venture capital funds or corporate
venture capital and has the primary aim to identify promising
investment opportunities for these investors. This accelerator type
resembles most of the original concepts of Y Combinator and
Techstars developed in the US. Its objective is to bridge the equity
gap between early-stage projects and investable businesses.
The deal-flow maker typically provides some form of seed fi-
nancing to the portfolio companies in exchange for equity. The
screening criteria in these programs tend to favor ventures that are
eligible for follow-on capital and have the ability to evolve in at-
tractive investment propositions. The mentors used in these ac-
celerators are often active business angels themselves, who play a
further role in follow-up investments. The director of Fintech In-
novation Lab described the mentors of deal-flow makers as
“investors in disguise”.
Deal-flow maker accelerators tend to select ventures, which
already have some proven track record or in some cases have al-
ready raised pre-seed finance. They hence focus on start-ups that
are in the later stages of development and often choose to spe-
cialize within a specific industry. By focusing on one specific sec-
tor, the accelerator management team can develop the necessary
sector-specific knowledge and expertise to identify and exploit the
economic potential of entrepreneurial teams.
4.2.3. The “welfare stimulator”
The “welfare stimulator”accelerator typically has government
agencies as a main stakeholder. The primary objective of this type
of accelerators is to stimulate start-up activity and foster economic
growth, either within a specific region or within a specific tech-
nological domain. For instance, the European Commission sup-
ports the establishment of accelerators within particular techno-
logical domains of its economic development program (i.e.
Knowledge and Innovation Communities or KICs).
The selection criteria and processes used in these accelerators
are oriented towards attracting companies that fit within the vi-
sion of welfare creation. For example, the Paris-based accelerator
Scientipôle Initiative promotes its program to unemployment
agencies in order to encourage unemployed entrepreneurs to ap-
ply to the accelerator. It focuses heavily on the potential for job
creation in its selection criteria.
Welfare stimulators typically select ventures in a very early
stage. Quite often a value proposition has not yet been developed.
As a consequence, the curricula and training programs provided by
welfare stimulators are most developed among the three types of
accelerators. Welfare stimulators typically organize training ses-
sions, workshops and practical learning –oriented events to help
the ventures develop their idea and value proposition. The accel-
erator’s mentors are closely involved with the portfolio companies
and provide hands-on guidance and advice. In some cases mentors
are consultants or business developers, who –often on a paid basis
–help to commercialize the technology or sell the product/service
idea.
However, for a lot of welfare stimulator accelerators, the busi-
ness model is rather unclear, as most public sponsors require some
form of revenue after an initial financing period. Although most
welfare stimulator accelerators present the typical investment
model as a potential, others experiment with other forms of rev-
enues such as tuition fees or registration fees for particular
training courses.
The above findings suggest that the accelerator design themes
are determined by the objectives of the affiliated shareholders
(respectively corporates, investors and government agencies). The
objectives of these shareholders; building a company ecosystem in
the case of corporates, identifying interesting investment oppor-
tunities in the case of investors and stimulating start-up activity
and economic development in the case of government agencies,
are translated into the primary objective of the accelerator and
represent the common theme orchestrating and connecting the
accelerator’s different design elements.
However, our data also point to the existence of hybrid accel-
erator types, which incorporate characteristics of two different
accelerator types. For examples, the London-based accelerator
Bethnal Green Ventures combines characteristics of the deal-flow
maker and welfare stimulator. The accelerator receives funding
from the UK Cabinet Office, Nominet Trust and Nesta and runs like
a traditional deal-flow maker accelerator in many aspects. It fo-
cuses on high-potential technology start-ups and invests up to
d15,000 in exchange for 6% equity. However, Bethnal Green Ven-
tures also has a strong social dimension. It is a strong advocate of
“Tech for Good”and exclusively focuses on companies that
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎ 9
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
leverage products and services for social good. In addition to fi-
nancial support it plays an important role in hosting meetings and
events in order to build a social community around the portfolio
companies and foster economic welfare creation.
5. Discussion and implications
This study extends previous incubation research by delineating
the accelerator model as a new generation incubation model, by
revealing the distinctive features of the accelerator model and
identifying the heterogeneity of accelerator strategies and opera-
tions. The extant incubation literature already identified a number
of descriptive characteristics of incubation models, resulting in a
variety of typologies and classifications, but, so far, failed to pro-
vide systematic evidence about whether these insights hold for
accelerators as well (Hackett and Dilts, 2004). Moreover, it lacks a
consistent theoretical framework to define and assess different
generation incubation models in order to account for the hetero-
geneity among incubation models and keep track of incubation
model evolution. This study addresses these gaps and thereby
provides important implications for both theory and practice.
5.1. Theoretical implications
Against a background of sparse research about accelerators, our
study has several implications for research on incubation models
in general and research into the accelerator model in particular.
First, we respond to the call in extant incubation research to
take the heterogeneity among incubation models into account by
delineating the accelerator model as a new generation incubation
model, by identifying its key design parameters and by shedding
light on the heterogeneity within the accelerator model. We show
that accelerators are different from other incubation models in five
aspects (program package, strategic focus, selection process,
funding structure and alumni relations) and highlight the objec-
tives of the accelerator’s shareholders as the main driver orches-
trating an accelerator’s activities. By identifying three different
groups of accelerators, we further contribute to the request from
incubation scholars to take different types of incubation models
and their specific features into account in order to assess perfor-
mance (Barbero et al., 2014;Mian, 1997).
Second, our results show that accelerator programs adopt dif-
ferent ways of structuring and running their programs, and that
this is largely determined by the objectives of their key share-
holders. Although most accelerator managers in our study men-
tioned Silicon Valley based accelerators Y Combinator and Tech-
stars as sources of inspiration, many of them do not adopt a pure
deal-flow maker model. We find two other types (the ecosystem
builder and the welfare stimulator) prevalent in Europe. The three
accelerator types differ in satisfying different shareholder needs
(respectively those of investors, corporates and public agencies).
As a consequence the deal-flow maker focuses heavily on men-
toring by serial entrepreneurs and business angels, who know how
to create legitimacy for follow-up investments. This is in line with
Kim and Wagman (2014), who suggest that accelerators act as
certification intermediaries, providing information and services
(e.g. screening practices and mentoring) valued by outside in-
vestors to help their portfolio ventures raise new capital. The
ecosystem builder is mainly focused upon helping ventures
through the complex decision making structures of corporate
companies. Instead of mentors, internal members of the corpo-
rates are used to support and guide the portfolio companies. Fi-
nally, the welfare stimulator tends to be more program-led by
providing intensive workshops and training sessions to help the
ventures find their way to first customers. With this finding, we
highlight shareholder objectives as important design parameters
to take into account, in addition to those of the portfolio compa-
nies participating in the accelerator. Previous research has argued
that the variety of incubation models is driven by the evolution of
portfolio companies’requirements and needs, which encourage
incubation mechanisms to differentiate the range of services that
they offer. However, our evidence leads us to argue that differ-
entiation between accelerators is driven by additional factors.
Specifically, from our qualitative evidence we theorize that dif-
ferences in the objectives of shareholders supporting or financing
the accelerator will lead to differences in the way accelerators run
their programs. Although portfolio companies’objectives do im-
pact the design of an incubation model (after all, changing port-
folio companies’needs gave rise to the accelerator model in the
first place), our study highlights the importance of other stake-
holder objectives, especially those stakeholders supporting and/or
financing the accelerator, to explain heterogeneity among different
accelerator model designs.
Third, by introducing a design lens to look at the accelerator
model, we contribute to recurrent requests in incubation research
to develop more theoretically grounded approaches to analyze
incubation activities (Bruneel et al., 2012;Hackett and Dilts, 2004).
We propose the activity system design perspective, highlighting
design elements and themes as important design parameters to
take into account, as an adequate theoretical lens to study in-
cubation models and their evolution. The design lens offers a
structured way to identify the key building blocks of the incuba-
tion model, enables classification of different incubation models,
as well as allows heterogeneity within the model to be taken into
account. Moreover, an additional advantage of this framework is
that it allows accounting for hybrid models. Within our sample we
note that two accelerators have hybrid models. Bethnal Green
Ventures has a clear welfare stimulation focus but nevertheless
copies the mentorship model typically present at the deal-flow
maker model, while Healthbox has a clear ecosystem building
focus but also provides some capital to its start-ups (see also Ta-
ble 3). The introduction of a design lens in incubation research
embodies rich possibilities for further theoretical development
and refinement. It not only gives researchers a concrete tool to
study incubation models and their evolution but also brings the
importance of design thinking, i.e. the design of an incubation
model is seen as a key decision in the creation of an incubation
entity, to the forefront of incubation research.
5.2. Managerial and policy implications
The accelerator design elements and themes identified in this
study can be used to position different accelerators within the
overall ecosystem. We suggest that initial advisors to early-stage
ventures (e.g. government support agencies; university student
and alumni entrepreneurship offices) should consider the different
accelerator design elements and themes in order to orient nascent
entrepreneurs towards particular types of accelerators that may
best meet their needs.
The diversity of accelerators we have identified also has im-
plications for policymakers in supporting different types of ac-
celerators and evaluating their role. Rather than evaluating the
effectiveness of an accelerator using using a fixed set of criteria,
there is a need to develop measures that take into account the
different objectives of different types of accelerators. Policymakers
typically have specific objectives, such as regional development
and employment. Taking these objectives into account, policy-
makers have to realize that the accelerators they finance might not
necessarily be profitable in the short or even medium term. The
ventures they invest in, the program they have to develop and
their strategic focus do not always allow this. The systematic
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎10
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
research evidence is sparse, but only deal-flow maker accelerators
in very dense ecosystems such as Silicon Valley appear to have a
proven business model. Unfortunately, we often see that policy-
makers expect welfare stimulators to have similar outputs as deal-
flow makers.
As accelerators have grown in popularity, many nascent en-
trepreneurs and organizations such as universities, companies and
regional development agencies feel attracted to the idea of starting
an accelerator. Universities see it as a way to promote student
entrepreneurship, companies as a way to tap into start-up in-
novation and talent, and development agencies as a way to create
employment. Examples of university-led accelerators include
“Beta Foundry”at Oxford University, InnovationRCA at the Royal
College of Art and the pre-accelerator “Imperial Create Lab”at
Imperial College, London. Our research shows that starting an
accelerator needs a clear vision and strategy, and a good fit be-
tween the different design parameters and the objective one
wants to achieve with the accelerator. Given the results so far, it
seems unlikely that accelerators will be profitable or even sus-
tainable without continued financial support for a number of
years. Although accelerators play an important role, the need for
this type of support needs to be legitimate. If not, the accelerator
initiatives will disappear as soon as the financial support for them
decreases.
Finally, our findings suggest that accelerators may help solve
some of the problems associated with previous generation in-
cubation models. Earlier, some incubation models have been ac-
cused of merely acting as life support and keeping tenants alive in
order to secure rent and fill their incubation space. As most ac-
celerators invest in their start-ups the accelerator model has an
added incentive to make sure that the selected start-ups survive
and scale. Accelerators are a way to shorten the journey of start-
ups, resulting in either quicker growth or quicker failure. However,
as some accelerators do allow alumni to remain in the space after
the program has ended, we have to take into account the potential
of creating adverse consequences if not time limited.
5.3. Limitations and future research
As all studies, this study is not without limitations. This final
section aims at outlining the particular limits of this study, which
provide interesting avenues for further research. First, the paper is
based on accelerators located in the three leading accelerator re-
gions in Europe: London, Paris and Berlin. These different Eur-
opean regions imply different contexts in which accelerators need
to function and be sustainable. However these three regions may
not be representative of all types of regions in Europe. As spatial
context may have an important influence on entrepreneurial and
innovation ecosystems (Levie et al., 2014), further research is
needed to test our findings in similar regions in other countries
and in different environments in general. Moreover, further re-
search is needed to examine the influences of policy, industry,
density and economic conditions on the configuration of different
accelerator types in a particular region.
Second, as accelerator programs develop, our framework,
highlighting the accelerator’s key design parameters, can serve as
a basis for more rigorous evaluations of accelerator performance
and can be used to define suitable success metrics in achieving
certain objectives. Subsequent analyses might also usefully ex-
amine the challenges faced by particular accelerators as they
evolve over time into different models, depending on the success
of their initial configuration.
Third, whilst beyond the scope of this paper, which has focused
on the accelerator as a unit of analysis, the study of the impact of
different accelerator types on their portfolio companies might be
an interesting avenue for further research as well. The approach
used by the accelerator is likely to have an impact on the en-
trepreneurial journey of start-ups and on the value added to them.
Further research on the differences between different accelerator
types and their impact on the entrepreneurial process would be
interesting, as this would enable identification of best practices
with the aim of implementing a customized acceleration strategy
to propel start-ups.
Finally, in order to truly gauge the effectiveness of different
models there is a need for studies that compare accelerated ven-
tures to a control group of non-accelerated ventures in order to
provide robust insights into the contribution of accelerators. Fur-
thermore, as decision makers perceive a focus on one sector or
technology as an interesting strategic option, assessment of dif-
ferences in effectiveness and value-added contributions to the
start-ups can improve our understanding of the possible benefits
of specialist versus generalist accelerators.
6. Conclusion
Accelerators play an important role in stimulating en-
trepreneurship. However, prior research has provided only limited
insight into their distinctive features and the heterogeneity of
their strategies and operations. Against a background of sparse
prior research, this study has produced several interesting results
about an accelerator’s key design parameters that have novel im-
plications for the incubation literature and practice. Obviously,
because the phenomenon is so new, uncertainty still exists about
the future success of accelerators. What is undeniable, though, is
the compelling economic logic of such organisations. We hope that
the findings of our study will open the way for further systematic
analyses of the processes and impacts of accelerator programs.
Acknowledgements
The special themed paper on Technology Business Incubation
Mechanisms would not have been possible without the insightful
feedback from two anonymous reviewers and the editors, for
which we are deeply grateful. The authors thank the innovation
charity NESTA and the Entrepreneurship Hub, Imperial College
London Business School, for their financial support.
References
Aernoudt, R., 2004. Incubators: tool for entrepreneurship? Small Bus. Econ. 23 (2),
127–135.
Amit, R., Zott, C., 2001. Value creation in e-business. Strateg. Manag. J. 22 (6-7),
493–520.
Amit, R., Zott, C., 2012. Creating Value Through Business Model Innovation. MIT
Sloan Manag. Rev. 53 (3).
Barbero, J.L., Casillas, J.C., Ramos, A., Guitar, S., 2012. Revisiting incubation perfor-
mance How incubator typology affects results. Technol. Forecasting Soc. Change
79 (5), 888–902.
Barbero, J.L., Casillas, J.C., Wright, M., Garcia, A.R., 2014. Do different types of in-
cubators produce different types of innovations? J. Technol. Transf. 39 (2),
151–168.
Becker, B., Gassmann, O., 2006. Corporate incubators: industrial R&D and what
universities can learn from them. J. Technol. Transf. 31 (4), 469–483.
Birdsall, M., Jones, C., Lee, C., Somerset, C., Takaki, S., 2013. Business accelerators:
The evolution of rapidly growing industry. University of Cambridge, Cambridge
(MBA Dissertation ad Judge Business School and Jesus College).
Bergek, A., Norrman, C., 2008. Incubator best practice: a framework. Technovation
28 (1), 20–28.
Bruneel, J., Ratinho, T., Clarysse, B., Groen, A., 2012. The evolution of business in-
cubators: comparing demand and supply of business incubation services across
different incubator generations. Technovation 32 (2), 110–121.
Carayannis, E.G., von Zedtwitz, M., 2005. Architecting gloCal (global-local), real-
virtual incubator networks (G-RVINs) as catalysts and accelerators of en-
trepreneurship in transitioning and developing economies: lessons learned and
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎ 11
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
best practices from current development and business incubation practices.
Technovation 25 (2), 95–110.
Christiansen, J., 2009. Copying Y Combinator, a framework for developing seed
accelerator programmes. University of Cambridge, Cambridge (MBA Disserta-
tion at Judge Business School and Jesus College).
Clarysse, B., Bruneel, J., 2007. Nurturing and growing innovative start-ups: the role
of policy as integrator. R&D Manag. 37 (2), 139–149 .
Clarysse, B., Wright, M., Lockett, A., Van de Velde, E., Vohora, A., 2005. Spinning out
new ventures: a typology of incubation strategies from European research in-
stitutions. J. Bus. Ventur. 20 (2), 183–216.
Cohen, S., Hochberg, Y.V., 2014. Accelerating startups: The Seed Accelerator Phe-
nomenon. Available at SSRN 2418000.
Dettwiler, P., Lindelof, P., Lofsten, H., 2006. Utility of location: a comparative survey
between small new technology-based firms located on and off Science Parks-
Implications for facilities management. Technovation 26 (4), 506–517.
Easterby-Smith, M., Thorpe, R., Holman, D., 1996. Using repertory grids in man-
agement. J. Eur. Ind. Train. 20 (3), 3–30.
Eisenhardt, K.M., 1989. Building theories from case-study research. Acad. Manag.
Rev. 14 (4), 532–550.
Eisenhardt, K.M., Graebner, M.E., 2007. Theory building from cases: opportunities
and challenges. Acad. Manag. J. 50 (1), 25–32.
George, G., Bock, A.J., 2011. The business model in practice and its implications for
entrepreneurship research. In: Entrepreneurship Theory Practice, 35 (1) pp. 83–
111.
Gioia, D.A., Price, K.N., Hamilton, A.L., Thomas, J.B., 2010. Forging an identity: an
insider-outsider study of processes involved in the formation of organizational
identity. Adm. Sci. Quat. 55 (1), 1–46.
Grimaldi, R., Grandi, A., 2005. Business incubators and new venture creation: an
assessment of incubating models. Technovation 25 (2), 111–121.
Hackett, S.M., Dilts, D.M., 2004. A systematic review of business incubation re-
search. J. Technol. Transf. 29 (1), 55–82.
Hill, S.A., Birkinshaw, J., 2014. Ambidexterity and survival in corporate venture
units. J. Manag. 40 (7), 1899–1931.
Huberman, A.M., Miles, M.B., 1983. Drawing valid meaning from qualitative data-
some techniques of data reduction and display. Qual. Quant. 17 (4), 281–339.
Isabelle, D.A., 2013. Key factors affecting a technology entrepreneur’s choice of
incubator or accelerator. Technol. Innov. Manag. Rev. 3 (2).
Kim, J.-H., Wagman, L., 2014. Portfolio size and information disclosure: An analysis
of startup accelerators. J. Corp. Financ. 29, 520–534.
Levie, J., Autio, E., Acs, Z., Hart, M., 2014. Global entrepreneurship and institutions:
an introduction. Small Bus. Econ. 42 (3), 437–444.
Mian, S.A., 1997. Assessing and managing the university technology business
incubator: an integrative framework. J. Bus. Ventur. 12 (4), 251–285.
Miles, M.B., Huberman, A.M., 1994. Qualitative data analysis: An expanded sour-
cebook, .. Sage
Miller, P., Bound, K., 2011. The Startup Factories: The rise of accelerator programmes
to support new technology ventures. NESTA.
Phan, P.H., Siegel, D.S., Wright, M., 2005. Science parks and incubators: observa-
tions, synthesis and future research. J. Bus. Venturing 20 (2), 165–182.
Salido, E., Sabás, M., Freixas., P. 2013. The Accelerator and Incubator Ecosystem in
Europe. In: Telefonica Europe.
Schwartz, M., 2013. A control group study of incubators’impact to promote firm
survival. J. Technol. Transf. 38 (3), 302–331.
Smilor, R., Gill, M., 1986. The New Business Incubator: Linking Talent, Technology,
Capital and Know-how. Lexington Books, Lexington, MA, p. 224.
Soetanto, D.P., Jack, S.L., 2013. Business incubators and the networks of technology-
based firms. J. Technol. Transf. 38 (4), 432–453.
Symon, G.E., Cassell, C.E., 1998. Qualitative Methods and Analysis in Organizational
Research: A Practical Guide. Sage Publications Ltd., California.
Szerb, L., Acs, Z., Autio, E., Ortega-Argilés, R., Komlósi, É., 2014. REDI: The Regional
Entrepreneurship and Development Index–Measuring regional entrepreneur-
ship: Downloaded from: ec.europa.eu/regional_policy/sources/docgener/stu-
dies/pdf/regional_ entrepreneurship_development_index.pdf.
Tan, F.B., Hunter, M.G., 2002. The repertory grid technique: a method for the study
of cognition in information systems. MIS Quat. 26 (1), 39–57.
Tracy, S.J., 2010. Qualitative quality: eight “Big-Tent”criteria for excellent qualita-
tive research. Qual. Inq. 16 (10), 837–851.
Van Looy, B., Debackere, K., Andries, P., 2003. Policies to stimulate regional in-
novation capabilities via university-industry collaboration: an analysis and an
assessment. R&D Manag. 33 (2), 209–229.
Vanderstraeten, J., Matthyssens, P., 2012. Service-based differentiation strategies for
business incubators: Exploring external and internal alignment. Technovation
32 (12), 656–670.
Wise, S., Valliere, D., 2014. The impact on management experience on the perfor-
mance of start-ups within accelerators. J. Priv. Equity 18 (1), 9–19.
Yin, R.K., 2013. Case Study Research: Design and Methods. Sage publications,
California.
Zott, C., Amit, R., 2007. Business model design and the performance of en-
trepreneurial firms. Organ. Sci. 18 (2), 181–199.
Zott, C., Amit, R., 2010. Business model design: an activity system perspective. Long.
Range Plan. 43 (2–3), 216–226.
Zott, C., Huy, Q.N., 2007. How entrepreneurs use symbolic management to acquire
resources. Adm. Sci. Quat. 52 (1), 70–105.
C. Pauwels et al. / Technovation ∎(∎∎∎∎)∎∎∎–∎∎∎12
Please cite this article as: Pauwels, C., et al., Understanding a new generation incubation model: The accelerator. Technovation (2015),
http://dx.doi.org/10.1016/j.technovation.2015.09.003i
