Conference PaperPDF Available

What is an Accelerator? The Case for a Pluralist Conceptualization

June 2020

June 2020

Conference: 2020 Academy of Management Annual Meeting

Authors:

Benjamin W. Walker

Victoria University of Wellington

Stephen Cummings

Victoria University of Wellington

Rebecca Downes

Victoria University of Wellington

Ruth Fischer-Smith

Victoria University of Wellington

Show all 5 authorsHide

The past decade has seen a marked rise in research on entrepreneurial accelerators. Efforts to develop a conceptual definition of the accelerator initially lagged behind this surge, but have recently begun to emerge in published articles. We explore this conceptual evolution as a case study of how scholars respond to new entrepreneurial phenomena. By first analyzing 48 academic definitions of the accelerator, we demonstrate an increasing convergence on Cohen and Hochberg's (2014) definition. By then analyzing practitioner understandings of the accelerator phenomenon using textual data from the websites of 73 accelerators, we find that practitioner understandings exhibit similarities to-but also some important differences from-the emerging academic conceptualization. We discuss the significance of this finding in light of the apparent convergence on a single academic definition of the accelerator, which, while conducive to efficient theory generation, might come at the expense of deep insight into the phenomenon. Mindful of this risk, and informed by our research findings, we make the case for a pluralist conceptualization by outlining four ways of viewing what accelerators can be: facilitators of new venture growth, vehicles for personal development, communities of practice, and strategic symbols. In light of the novel and continuously evolving nature of many entrepreneurial phenomena, we conclude by discussing the theoretical benefits of pluralist conceptualizations for the field of entrepreneurship as a whole.

Relative frequency of definitional elements across accelerator websites (n=73)

…

Elements of Academic Definitions of the Accelerator

…

Frequency of definitional elements across accelerator websites

…

Figures - uploaded by Benjamin W. Walker

Content may be subject to copyright.

Content uploaded by Benjamin W. Walker

Content may be subject to copyright.

What is an Accelerator? The Case for a Pluralist Conceptualization

Working Paper @ June 2020

Benjamin W. Walker

ben.walker@vuw.ac.nz

Victoria University of Wellington

New Zealand

Stephen Cummings

Victoria University of Wellington

New Zealand

Rebecca Downes

Victoria University of Wellington

New Zealand

Ruth Fischer-Smith

Land Information New Zealand

Jesse Pirini

Victoria University of Wellington

New Zealand

ABSTRACT

The past decade has seen a marked rise in research on entrepreneurial accelerators. Efforts to

develop a conceptual definition of the accelerator initially lagged behind this surge, but have

recently begun to emerge in published articles. We explore this conceptual evolution as a case

study of how scholars respond to new entrepreneurial phenomena. By first analyzing 48

academic definitions of the accelerator, we demonstrate an increasing convergence on Cohen

and Hochberg’s (2014) definition. By then analyzing practitioner understandings of the

accelerator phenomenon using textual data from the websites of 73 accelerators, we find that

practitioner understandings exhibit similarities to — but also some important differences from

— the emerging academic conceptualization. We discuss the significance of this finding in

light of the apparent convergence on a single academic definition of the accelerator, which,

while conducive to efficient theory generation, might come at the expense of deep insight into

the phenomenon. Mindful of this risk, and informed by our research findings, we make the

case for a pluralist conceptualization by outlining four ways of viewing what accelerators can

be: facilitators of new venture growth, vehicles for personal development, communities of

practice, and strategic symbols. In light of the novel and continuously evolving nature of

many entrepreneurial phenomena, we conclude by discussing the theoretical benefits of

pluralist conceptualizations for the field of entrepreneurship as a whole.

Keywords:

Incubators and accelerators; Conceptual theory; Critical perspectives on accelerators

What is an Accelerator? The Case for a Pluralist Conceptualization

Within entrepreneurial ecosystems, the accelerator (sometimes referred to as “seed

accelerator” or “startup accelerator”) is increasingly being adopted as a model for cultivating

ventures. In the United States, the number of active accelerators increased tenfold between

2006 and 2016 (Clifford, 2016), while the database Pitchbook listed almost 900 accelerators

in operation worldwide in 2015 (Widjaja & Tom, 2015). This proliferation has inevitably led

scholars to try and better understand the phenomenon, resulting in the emergence of a

promising new literature on accelerators (e.g., Bliemel, Flores, De Klerk, & Miles, 2019;

Cohen, Bingham, & Hallen, in press; Cohen & Hochberg, 2014; Pauwels, Clarysse, Wright,

& Van Hove, 2016; Walshook, 2013).

Within this burgeoning literature, though, scholars are only just starting to explore the

question of what an accelerator actually is from a conceptual standpoint. This is not to say

that academics have failed to define the accelerator, but rather that existing definitions are

often put forward in an ad-hoc manner (e.g., “For the purposes of this paper, an accelerator is

defined as…”), or with limited consideration of the broader diversity in the structures and

aims of existing accelerators around the world. Thus, while the definitions scholars have

proposed are often appropriate for the specific article they appear in, or study they relate to,

they are arguably less suited to being field-level definitions that guide the collective research

effort and enable comparative research moving forward.

Such an ad hoc approach to conceptualizing the accelerator is certainly

understandable, however, given that the phenomenon itself continues to evolve. The first

major accelerators, such as Y Combinator and Techstars, launched in the mid-2000s with the

primary aim of cultivating successful new companies, in a conventional for-profit sense.

Since then, myriad new accelerators have emerged that vary — sometimes significantly — in

both their aims and approach, including accelerators focused on not-for-profit organisations

and social enterprises (e.g., Casanovas & Bruno, 2013), and corporate accelerators that aim to

cultivate innovation within existing organisations (e.g., Richter, Jackson, & Schildhauer,

2018).

Many scholars have emphasized the importance of clear and accurate construct

specification (e.g., Johnson, Rosen, Chang, Djurdjevic, & Taing, 2012; Law, Wong, &

Mobley, 1998; Podsakoff, MacKenzie, & Podsakoff, 2016; Suddaby, 2010): it is challenging

(if not impossible) to develop a coherent body of knowledge on a phenomenon when

insufficient attention has been devoted to delineating what the phenomenon actually is. It is

apparent, therefore, that agreement amongst scholars on the meaning of constructs is central

to developing theory. Motivated by these arguments, and our view of the accelerator as a “live

case” where the quest for definition is being played out in real time, we explore the

conceptual evolution of accelerators as a case study of how scholars respond to new

phenomena and converge on definitions, and what the benefits and costs of this process may

be. Consequently, the questions guiding this research were as follows: 1) what are the central

elements of both scholarly and practitioner conceptualizations of the accelerator, 2) which of

these elements are (not) shared across both groups’ conceptualizations, and 3) what are the

implications of the answers to questions 1) and 2) for enhancing theory on accelerators

moving forward?

To explore these questions, we adopted a two-step research approach, which we report

in detail in this paper. In brief, we first assessed the emerging state of the academic

conceptualization of the accelerator by identifying the most common elements of 48 scholarly

definitions, and how these definitions have changed over time in which elements are

emphasized, leading to a recent convergence on one specific definition. Second, to ascertain

the validity of this conceptual evolution and convergence, we compared it with an analysis of

practitioner perspectives on accelerators, based on data collected from the websites of 73

(self-proclaimed) accelerators from around the world. Through this comparison, we found

that while there are some areas of alignment between how academics and practitioners

understand the accelerator, there are also areas of misalignment, suggesting the need for

further consideration of how we as scholars conceptualize accelerators.

Exploring the diversity of practitioner perspectives on the accelerator also led us to

consider what was being gained and lost as the definitional consensus amongst scholars

increased. The gains are relatively well understood. As we have indicated above, construct

clarity allows communities of scholars to agree on the fundamental nature of the objects they

study, which in-turn facilitates more efficient theoretical development (e.g., Locke, 2012;

Suddaby, 2010). However, in light of the findings of the present research, we also see value in

allowing space for multiple perspectives on phenomena, including the accelerator. While

some see a lack of construct clarity as a problem to be “fixed”, we advocate that variation in

conceptualizations of phenomena (in this case, the accelerator) can sometimes be

advantageous for encouraging innovation in research and developing a richer body of

knowledge on a phenomenon (e.g., Morgan, 2006; Örtenblad, Putnam, & Trehan, 2016) —

particularly when in practice the phenomenon is novel and continuously evolving.

In light of the above considerations, we conclude by discussing the value of

approaching phenomena from multiple conceptual perspectives in the field of

entrepreneurship, where many focal phenomena are both novel and changing rapidly over

short timeframes. To stimulate and provide a guiding framework for such efforts with regard

to accelerators specifically, we conclude by presenting four conceptual lenses through which

accelerators can be conceptualized, defined, and studied. We do so in the belief that this

pluralist approach to conceptualization will, in the long-term, yield a deeper, more nuanced

understanding of accelerators and their role in communities, economies, and society at large.

PART #1: ACADEMIC CONCEPTUALIZATIONS OF THE ACCELERATOR

In light of our desire to progress towards a comprehensive understanding of the accelerator, it

was important to first gain a detailed understanding of how scholars presently conceptualize

the accelerator, and how this conceptualization might have evolved since research on

accelerators began. To this end, we report here the findings of our analysis of 48 published

academic definitions of the accelerator — all of which have been articulated in the past 8

years. Our analysis was both descriptive and developmental in its focus: we sought to a)

identify the most common elements of existing academic definitions of the accelerator, and b)

assess how academic definitions have changed over time in which of these definitional

elements they emphasize.

Research Approach

Given our interest in understanding the key elements of scholarly conceptualizations (and

their evolution over time), our data was qualitative in nature — specifically written definitions

of the accelerator concept appearing in academic articles. Our analytic approach involved a

mix of both quantitative and interpretive content analysis (Patton, 2002), meaning we sought

to uncover common broad patterns in the definitions using both mathematical techniques and

data-informed judgment. We adopted this hybrid approach because it provides a desirable mix

of rigor, in that the use of quantitative techniques assured us of a basic degree of consistency,

and flexibility, in that we were able to exercise judgment where we noticed interesting

patterns across the definitions that were not necessarily captured by quantitative techniques.

Collection of definitions. We collected academic literature on accelerators

(specifically journal articles, conference papers, online pre-prints, and working papers) by

surveying citations in recently published articles on accelerators (e.g., Cohen et al., in press;

Cohen, Fehder, Hochberg, & Murray, 2019; Hallen et al., in press), and conducting searches

using the Web of Science and ProQuest databases. The second author searched the databases

using the term “accelerator” combined with any of the terms “innovation”, “entrepreneurship”

or “entrepreneur” to limit results to research related to venture acceleration (rather than, for

example, photon acceleration). Results from the database searches were cross-checked against

identical searches on Google Scholar, which allowed us to identify newer work citing seminal

literature. Through this approach, we identified a total of 85 academic papers on accelerators,

all of which were published after 2000.

The second author then sought to identify and collect definitions of accelerators from

each paper. First, the paper was searched for the words “define” and “definition” to identify

any explicit definitions. Where no explicit definition was apparent, the abstract, introduction,

and literature review/theoretical background sections were all inspected for more implicit

definitions. Some research, particularly early work prior to the existence of widely accepted

definitions, provided lengthy descriptions of accelerators rather than a concise definition.

These descriptions would typically outline features, goals, outcomes, and examples of

accelerators, often based on prior research, and would span several sentences or several

paragraphs. While these descriptions outline what the author(s) consider to be an accelerator,

we did not deem them to be definitions for our purposes. In the case of empirical papers, the

method section was also inspected to determine any criteria used to identify accelerators as

part of sample selection. Where no definition was provided but selection criteria (i.e., for

building a sample of accelerators in a study) were described, we recorded these criteria as a

definition. Where both a definition and selection criteria were provided, both were recorded.

Finally, the text of each paper was checked using a search for the word “accelerator” to

ensure that any definitions elsewhere in the text had not been identified. Through this

approach, we identified a total of 48 definitions of the accelerator appearing in academic

literature. Note that these were not necessarily unique definitions – indeed as we show

shortly, many of them were restatements or paraphrased versions of earlier definitions.

Analysis of definitions. To gain a high-level understanding of the most common

elements of the definitions we collected, we imported the list of definitions into the qualitative

data analysis software NVivo and executed the word count function, limited to the 30 most

common words. After removing irrelevant words from the list (e.g., “the”, “that”, “refers”),

the first author used this initial list of frequently-appearing words as a coding template for

manual analysis of the definitions. The first author read each definition, and recorded which

words appeared in the definition. In the process, the first author also considered whether there

were any potentially significant terms (in terms of understanding the accelerator

phenomenon) that appeared in the definitions but were not represented in the initial list

produced by the word count, and updated the list accordingly. In this way, our research

approach was abductive in nature (Reichertz, 2007): while we used our list of frequently-

appearing words as an initial conceptual guide, we refined and added to this conceptual guide

as needed based on close reading and interpretation of each definition.

Through the process described above, we developed a list of 12 elements that were

common (to varying degrees) across the definitions we collected (see the “definitional

elements” column of Table 1). As the analysis progressed, we also collectively observed that

these definitional elements could be broadly grouped into two meaningful categories –

elements that define the accelerator based on its structure, and elements that define the

accelerator based on what it supposedly offers entrepreneurs and/or their ventures. Finally, we

were also able to assess the relative commonality of each element across all of the 48

definitions, and by taking into account publication year, the extent to which the frequency of

each definitional element had increased/decreased over time. Because of the small number of

definitions published in the early years of research on accelerators (n=1 for 2012, n=3 for

2013), and the relatively slow pace of academic publication generally, we grouped definitions

into two-yearly rather than yearly intervals for our analysis.

- - - - - - - - - - - - - -

Insert Table 1 here

- - - - - - - - - - - - - -

Findings

In terms of the relative frequency of each of the definitional elements across the 48 definitions

we collected, our findings are summarised in the right-most column of Table 1, and visually

in Figure 1 below. Of the specific definitional elements we identified, “limited duration”

(74% of definitions), “cohort-based” (70%), “education” (70%), “mentorship/coaching”

(60%), and “concluding event” (45%) were the five most common elements across the

definitions. After these five elements, there is a rather steep drop-off, with the remaining 7

elements being relatively equal in frequency, except for “physical space” (11%), which was

the least frequently appearing definitional element by a considerable degree. More broadly, it

is interesting to note that 3 of the 5 most frequently appearing elements (“limited duration”,

“cohort-based”, “concluding event”) concern the structure of accelerators, while the

remaining 2 (“education”, “mentorship/coaching”) concern the supposed offerings of

accelerators. It is also notable that both of the offering elements concern learning or skill

development: from an academic standpoint, it seems that the most common framing of the

function of the accelerator concerns their potential to develop the capabilities of

entrepreneurs, as opposed to the less personal, more direct approach of directly developing

ventures (e.g., via funding or strategic guidance). This point is further supported by the fact

that the element “venture acceleration”, which refers to accelerators’ assisting ventures to

scale and grow (or at least “fail fast”), was one of the least common elements (19%) across

the definitions.

- - - - - - - - - - - - - - - - - - -

Insert Figures 1 and 2 here

- - - - - - - - - - - - - - - - - - -

Figure 2 helps to situate the above findings in historical context, and demonstrates

some interesting shifts in what academic definitions of the accelerator have emphasized since

2012. Across the definitions published in 2014/2015, for instance, the most common

definitional element was “physical space”, with 80% of definitions claiming that provision of

a workspace to nascent entrepreneurs was a defining feature of the accelerator — something

that was far less common in the 2016/2017 period, with only 13% of definitions making

reference to the physical space criterion. Similarly, “direct investment” by accelerators in

accepted ventures was at one point a fairly common definitional element, with 42% of

definitions in the 2014/2015 period referencing this element. However, the direct investment

criterion has steadily decreased since then, with only 24% of definitions making reference to

direct investment in the 2018/2019 period.

What factor(s) are driving the changes in definitional emphasis charted in Figure 2?

Based on further analysis of citations appearing in the definitions, it seems the primary reason

for this shift in definitional emphasis is that the literature has in the past few years converged

around Cohen and Hochberg’s definition of the accelerator: “a fixed-term, cohort-based

program, including mentorship and educational components, that culminates in a public pitch

event or demo-day” (2014: 4). This inference is supported by the fact that of the 48

definitions we collected, 13 (27%) directly cite Cohen and Hochberg (2014), easily making it

the single most cited definition across the definitions we collected. Interestingly, this

definition appears in an unpublished pre-print – not a peer-reviewed journal article.

Part #1 Summary

Our findings show that since the emergence of the first accelerators over a decade ago,

academic definitions of the accelerator have varied in which aspects of the phenomenon are

emphasized, but since 2017, have begun to converge around Cohen and Hochberg’s (2014)

definition. A crucial follow-up question, though, is to what extent does this particular

conceptualization align with the structures and offerings of real-world accelerators: is it

appropriate to view accelerators as fundamentally defined by their cohort-based, limited

duration structure, and as primarily focused on offering education and mentorship to aspiring

entrepreneurs? In Part #2 of our study, we sought to explore these questions by analysing the

online claims of real-world accelerators.

PART #2: PRACTITIONER CONCEPTUALIZATIONS OF THE ACCELERATOR

After assessing how scholars have conceptualized the accelerator, we sought to uncover

current understandings of the phenomenon amongst those who actually establish and operate

them, with the ultimate aim of identifying the similarities and differences between the two

groups’ perspectives. We achieved this by analysing textual data collected from 73

accelerators across 8 countries. Below, we further detail our research approach, and then

report our findings with reference to relevant patterns we observed in our analysis of

academic definitions.

Research Approach

Sample. In an effort to build a (roughly) globally representative sample of accelerators, we

identified 10 of the most prominent accelerators headquartered in each of the following 8

countries — Finland, Ireland, Israel, Kenya, New Zealand, Singapore, USA, and the UK. To

identify which 10 accelerators to include from each country, we first adopted an exploratory

approach, consulting various “top accelerator”-style articles related to each country. We then

experimented with different Google search terms (e.g., “accelerator Israel”, “start-up

accelerator Israel”) to identify which search term consistently produced the most coherent list

of results across the 8 countries. By “coherent” list, we mean that the first 10 accelerators

returned by the search term seemed to correspond reasonably closely with what the “top

accelerator”-style articles we consulted noted as the most prominent accelerators in the

country in question. We concluded that searching for “accelerator program” + “country

name” generally yielded the most coherent list of accelerator websites in each country. Thus,

we completed this search for each of the 8 countries listed above, leading us to identify a total

of 80 accelerator websites from which we intended to collect data. During the data collection

process, 7 of these websites (2 from Finland, 3 from Singapore, 1 from Israel, and 1 from

Kenya) went offline and became inaccessible, presumably because the associated accelerator

stopped running. Because we had incomplete data for these 7 websites, we omitted them from

our sample, leaving us with a final sample of 73 accelerator websites. Table 2 lists the name

and country of each accelerator included in our final sample.

- - - - - - - - - - - - - - -

Insert Table 2 here

- - - - - - - - - - - - - - -

Data collection. As each accelerator website had a substantial amount of potentially relevant

data, we needed to identify a means of structuring our data collection – at least in the initial

stages. To this end, we used the list of definitional elements that emerged from our analysis of

academic definitions (see Table 1) to focus our data collection. Thus, the second author (or in

some cases, a research assistant) read each section of each website in an effort to profile it

according to our initial list of definitional elements. As an example, the following statement

from the Frequently Asked Questions section of one accelerator website was coded as

reflecting the “concluding event” criterion:

“Q. Will there be a demo day?

A. Yes. At the end of the program you will present your company to a selected

audience of investors, corporate partners and press.”

We only recorded whether or not each website made a claim reflecting each of the

definitional elements at least once (i.e., we did not record repeated references, and thus, the

extent to which each website seemingly emphasized each definitional element). In addition, as

with our analysis of the academic definitions, we adopted an abductive approach, treating our

initial framework as open to revision in light of new and surprising data. Thus, as the process

of profiling each website unfolded, we noted any interesting pieces of data from the websites

that reflected an aspect of accelerators that did not appear in the academic definitional

elements list, and revised the list accordingly where a particular element emerged prominently

across the websites.

It is important to acknowledge at this stage that the material contained on accelerators’

websites is often promotional in content and language. Thus, the data we collected potentially

paints an overly positive picture of how accelerators operate and what they deliver. It is also

important to point out that the website data represents the views of those who establish and

operate accelerators, not necessarily the entrepreneurs who participate in them, or other

accelerator stakeholders (e.g., investors, policymakers). At the same time, we note that

through our efforts in Part #1, we found that the structure and offerings of accelerators

emerged as the two high-level mediums through which scholars tend to define accelerators.

Websites provide a wealth of information about both the structure and offerings of

accelerators, and therefore represent a data source that was both appropriate to our research

aims and straightforward to access. Nevertheless, the aforementioned idiosyncrasies of our

data should be borne in mind when interpreting our findings.

Data analysis. Through the approach described above, we were able to assess how

frequently each of the defining elements of accelerators identified in Part #1 (i.e., across

academic definitions) appeared across the sample of accelerator websites – both in absolute

terms and as a percentage of the 73 websites we analyzed. Importantly, we were also able to

identify aspects of accelerators that were commonly claimed by accelerators on their

websites, but not present in the academic definitions we reviewed in Part #1. In this way, we

were able to compare and contrast which aspects of accelerators were emphasized in

practitioner perspectives (using accelerator website data as a proxy) vis-à-vis academic

conceptualizations (using written definitions as a proxy), and also which aspects of

accelerators had been overlooked in academic definitions but commonly noted by accelerators

on their websites.

Part #2 Findings

Since we used the list of elements that emerged from our analysis of academic definitions to

guide what data we collected from accelerator websites, we present our findings in two main

sections – the first concerning accelerators’ claims about structure, and the second concerning

claims about offerings. Within each section, we also discuss similarities to and differences

from those elements that were present (and indeed most prevalent) in academic definitions of

the accelerator. Our findings are summarised numerically in Table 3 and visually in Figure 3.

For ease of comparison, in Figures 4 and 5 we display the relative frequency of structural

elements and offerings, respectively, across accelerator websites vs. academic definitions of

the accelerator.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Insert Table 3, and Figures 3, 4, and 5 here

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Structure of accelerators. Our findings regarding accelerators’ claims about their structure

suggest areas of both commonality and divergence from those structural elements emphasized

in academic definitions. In terms of commonalities, 91% of the accelerators whose websites

we consulted described themselves as limited duration programmes, while 87% described

themselves as cohort-based programmes. Recall that 70% and 74% of academic definitions

listed limited duration and cohort-based as defining features of the accelerator, respectively –

a consensus that certainly appears justified given the high volume of accelerators in our

sample that claimed such features.

In terms of areas of divergence, we found that 80% of the accelerators whose websites

we assessed noted some kind of significant concluding event (e.g., graduation, “demo day”,

pitch evening) as an aspect of their programme – a contrast with the fact that only 45% of

academic definitions included the concluding event criterion. Similarly, we observed that

85% of the accelerator websites we assessed described the accelerator as having some kind of

competitive application process – an intriguing finding in light of the fact that only 19% of

academic definitions made reference to an application process as a defining feature of

accelerators. These findings suggest that, from a scholarly standpoint, it may be appropriate to

give more consideration to the concluding event and application process criteria as defining

features of the accelerator. Finally, we observed that 75% of the accelerators in our sample

described themselves as having an exclusively team-based model (as opposed to a non-

exclusive model that allowed teams, duos, or single founders to participate), a contrast with

the fact that only 19% of the academic definitions made reference to the team-based nature of

accelerators as a defining feature.

We also inductively identified 4 categories of structural elements that did not emerge

from our analysis of academic definitions, but which were apparent across at least 20% of the

accelerator websites we assessed. The first of these we termed customizability, which

captured those claims by accelerators that they would adapt the nature of their program to suit

accepted ventures’ specific needs. 20% of the accelerator websites we assessed made

customizability claims, the following excerpt from the HATCH accelerator (Ireland) website

being illustrative of such claims:

“HATCH is a mentor-driven 15-week program and is tailored to meet the

requirements of the teams selected for each cohort.”

The second structural element we developed we termed time demands: 68% of the

accelerators whose websites we assessed claimed to offer a full-time programme, 15%

claimed to offer a part-time programme, and the remaining 17% provided no clear

information about the time demands of their programme. The third structural element we

developed we termed in-person, which captured whether or not the accelerator required

ventures to be physically present on their premises for a substantial portion of the programme.

71% of the accelerator websites we assessed stated an in-person requirement. The fourth

structural element we termed fee-based, which captured statements on websites indicating that

the accelerator’s business model relied on charging ventures a financial fee for participation

in the programme. 21% of the accelerator websites we assessed explicitly stated they were

fee-based.

Offerings of accelerators. As with the structural elements, when it came to the

offerings of accelerators, we also found areas of commonality and divergence between the

claims of accelerators on their websites and academic definitions. At a broad level, we found

that all of the 6 offering categories identified across the academic definitions in Part #1 also

appeared across websites (though not necessarily to equivalent degrees), and we also

inductively identified 6 new offering categories that were prevalent (to varying degrees)

across the accelerator websites but which did not emerge from our analysis of academic

definitions.

Beginning with the offering categories that emerged from our analysis of academic

definitions, we found that claims about education were virtually as frequent across accelerator

websites (72%) relative to academic definitions (70%). This finding suggests that the

emphasis on education as a defining offering of accelerators in the academic literature is

indeed justified. Similarly, we found that claims regarding mentorship/coaching were very

common across accelerator websites (81%), even more so than across academic definitions

(60%) — a finding that suggests scholars’ emphasis on mentorship/coaching as a defining

offering of accelerators is also justified. In terms of areas of divergence, certainly the most

notable concerned networking: 96% of accelerator websites made at least one claim regarding

provision of networking opportunities to ventures and/or entrepreneurs, compared with only

21% of academic definitions that included networking as a defining feature of the accelerator.

In addition, we observed much diversity in what types of networking accelerators claimed to

offer entrepreneurs: some claimed to offer access to experts or industry partners (e.g.,

suppliers, manufacturers), while others claimed to offer access to investors, alumni of the

accelerator itself, or international partners who could assist with overseas expansion. Such

findings indicate that real-world accelerators view access to other people, groups, or

organizations as a central offering, suggesting that the academic conceptualization of the

accelerator would do well to give greater weight to the networking criterion. Finally, we also

found that mentions of providing funding opportunities and physical space for entrepreneurs

to work in were more common across accelerator websites (57% and 52%, respectively)

versus academic definitions (21% and 11%, respectively).

We also inductively developed 6 new categories of offering claims that we noticed

were prevalent (to varying degrees) across accelerator websites, but which did not emerge

from our analysis of academic definitions. The first of these categories, market confirmation,

which appeared across 40% of the websites, we created to capture those claims by

accelerators that they could assist ventures in confirming there is a market for their product, or

that ventures were targeting the correct market/sector with their business model. The second

new category of offering claims, exposure, which appeared across 51% of the websites, we

created to capture those claims by accelerators that they could connect ventures with potential

customers for their product, and/or more generally increased awareness of the product

amongst the target market. The third new category of offering claims, non-financial

resources, which appeared across 41% of the websites, captured those claims by accelerators

that they could offer ventures access to non-financial resources (e.g., technology, R&D

laboratories, or Amazon Web Service credits). The fourth new category of offering claims,

business model development, appeared across 39% of the websites and captured accelerators’

claims to directly assist ventures with developing and/or refining their business model/plan.

The fifth new category of offering claims, product design/development, appeared across 25%

of the accelerator websites and captured claims by accelerators to assist ventures with

designing and/or developing their products

. The final new category of offering claims, which

was noticeably less technical and business-oriented than the others we developed, we termed

moral support, which captured those claims by accelerators to be enthusiastic and emotionally

supportive of ventures and/or entrepreneurs. These claims seemed to indicate that the

accelerator would be personally and emotionally invested in the venture's success (beyond

just financial outcomes). The following excerpt from the xEdu (Finland) accelerator website

is prototypical of such moral support claims, which we identified across 17% of the websites:

“If you get accepted to xEdu acceleration program it means that we believe in your

success. We believe that your EdTech product can reach the stars and our first

mission is to show you the way there. Let these planets of our expertise be your

guideline that will bring you to international success.”

Part #2 Summary

We explored practitioner conceptualizations of the accelerator by categorising the online

claims of 73 accelerators across 8 countries into 21 elements, and assessing the frequency of

each of these elements across accelerator websites. Consistent with the findings of our

analysis of academic definitions in Part #1, we found that practitioners also largely view

accelerators as having a limited duration, cohort-based structure. We also found that, like

academics, practitioners view education and mentorship as primary offerings of accelerators.

We also found differences — in terms of those elements most commonly emphasized

— between scholarly and practitioner conceptualizations: practitioners as a collective seem to

place significantly more emphasis than academics on networking, physical space, and funding

opportunities as outcomes of accelerators, and concluding events, application processes, and a

team-based model as features of accelerators’ structure.

We use the term “product” in its broadest sense here to refer to the central output of ventures (e.g., mobile

application, social service or initiative, physical good)

Perhaps the most interesting finding arising from Part #2 of our study, however, is the

fact that we inductively identified 6 new types of offering claims (moral support, product

design/development, business model refinement, market confirmation, non-financial

resources, exposure), beyond those 6 that emerged from our analysis of academic definitions.

This finding suggests that — at least from the perspective of those responsible for

establishing and operating accelerators — accelerators may deliver a greater range of

outcomes than presently captured by the academic conceptualization of the accelerator. It also

raises the question of whether attempting to specify all of the specific offerings of accelerators

in a definition is desirable, let alone possible: it may be more parsimonious to define the aims

of accelerators in more abstract terms (i.e., to facilitate the development of ventures and/or the

entrepreneurs who found them).

DISCUSSION: AN ARGUMENT FOR A PLURALIST CONCEPTUALIZATION OF

THE ACCELERATOR

We undertook a two-part case study aimed at understanding and comparing academic and

practitioner perspectives on the accelerator phenomenon. We did so as a means of considering

how we might best define this rapidly-emerging phenomenon. Part #1 of our study analysed

the evolution of academic conceptualizations (as represented in published definitions) of

accelerators over time. Part #2 analysed practitioner conceptualizations of accelerators based

on data collected from the websites of various accelerators around the world, with the aim of

comparing and contrasting this practitioner perspective with the academic conceptualization

canvassed in Part #1. The findings of our studies indicate that academic definitions and

practitioner perspectives on the accelerator have shared areas of emphasis, but also differ in

substantial ways. This raises a question that guides our discussion in this section: how exactly

should scholars interested in studying accelerators proceed, in terms of how we conceptualize

the phenomenon?

First, we would start from the point that the definition of accelerators proffered by

Cohen and Hochberg (2014), and which scholars are increasingly subscribing to (see Part #1

findings), is by no means a poor definition. Indeed several elements of this definition (i.e.,

limited duration, cohort-based, education, mentorship/coaching) are central elements shared

with the practitioner perspectives in Part #2 of our study. Yet given that we also identified

other elements of the practitioner perspective not represented in Cohen and Hochberg’s

(2014) definition, or even the academic literature on accelerators at large, there remains work

to do if our aim is to develop a truly comprehensive conceptualization of the accelerator.

Perhaps the most obvious option at this point would be to seek to develop a “new-and-

improved” definition of the accelerator — one that encapsulates the key elements of both

academic and practitioner perspectives reflected in our findings. As mentioned at the outset of

this article, consensus amongst scholars on definitions of phenomena has frequently been

noted as central to theoretical progress in scientific fields of inquiry (Locke, 2012; Suddaby,

2010). In the case of emergent phenomena like the accelerator, however, there are two

considerations that we argue makes convergence on a single, all-encompassing definition a

potentially undesirable aim in terms of theoretical comprehensiveness.

First, given the vast array of structural elements and offerings of accelerators that

emerged in our study of accelerator websites (see Part #2 findings), it seems reasonable to

conclude that practitioners — or at least those who establish and operate accelerators — are

presently answering the questions of what exactly accelerators are and do in a host of

different ways. Thus, it seems unlikely that any one definition of the accelerator would be

able to parsimoniously account for all or even most of such variation. Second, it must be

remembered that accelerators are ultimately socially constructed phenomena, which

inevitably change over time as a result of broader structural changes in values, culture,

politics, and society (Berger & Luckmann, 1966; Charmaz, 2014). Thus, even if one were

able to develop a single definition of the accelerator that neatly captured all of the variation in

conceptualizations that presently exists amongst both scholars and practitioners, there is no

guarantee that this definition would be appropriate into the future. Indeed based on what we

have personally observed in the past three years of researching accelerators, we suspect that

the variety of perspectives on what accelerators actually are amongst practitioners will only

grow in response to new social, cultural, and commercial issues that arise.

Thus, if we accept, ontologically, that accelerators are socially constructed phenomena

that can be conceptualized in a variety of ways, then it follows, epistemologically and

methodologically, that the definitions and research approaches scholars adopt to study them

should also vary (Kuhn, 1962; Foucault, 1970). Such onto-epistemological considerations

have exercised the more mature fields of management research (e.g., organizational behavior,

organizational theory) for some time (e.g., Calás & Smircich, 1999; Gioia & Pitre, 1990), and

we believe there is theoretical value to be gained from taking such considerations seriously in

the field of entrepreneurship generally, and also in the specific “live case” of the accelerator,

where the phenomenon is emerging at the same time as scholars seek to define it.

Even considering a few simple models of ontology and epistemology can be

insightful. A popular model like Burrell and Morgan’s Sociological Paradigms in

Organizational Analysis (1979), for example, is instructive in highlighting the potential

cost(s) of convergence on a single comprehensive definition. If we view a phenomenon from

a functionalist standpoint (i.e., one which resembles the conventional positivist approach of

the natural sciences), then a single, widely-accepted definition will be understood as a

necessary condition for theoretical progress. Closing off alternatives is therefore an essential

part of progressing toward a true understanding of the object in question. But, if we value

growth, change and innovation in research; if we believe that capturing reality and all of its

complexity is more important than scientific ‘efficiency’; or if we believe it is too early to

“close the gate” on defining what exactly an object is or must be, and that there is much to be

gained by letting innovation in practice and in our understanding unfold; then we must

recognize that a comprehensive unitary definition of a phenomenon is not necessarily a means

to our desired ends, at least when the phenomenon itself is still taking shape.

Mindful of the above arguments and of the findings we have reported in this

manuscript, we argue that at this early stage, rather than developing and promoting consensus

on a single definition of the accelerator, the quality and comprehensiveness of theorizing

might be better facilitated by encouraging researchers to approach the accelerator from

multiple conceptual perspectives: each of which makes different assumptions about the

fundamental purpose of the object under study (e.g., Morgan, 2006; Örtenblad et al., 2016).

As Gioia and Pitre observe, there is often value in framing conceptualization and theory

building in management studies “not as a search for the truth, but as more of a search for

comprehensiveness stemming from different worldviews” (1990: 587, emphasis in original).

Motivated by this view, and informed by our findings regarding the central elements

of academic definitions and practitioner understandings of the accelerator, we seek here to

spark a conversation about how a pluralist conceptualization of the accelerator (i.e., a

research agenda informed not by one but multiple definitions) could work. We do so by

outlining four perspectives on what accelerators can be, each of which encapsulates some, but

importantly not all, of the elements surfaced by our analysis of academic and practitioner

conceptualizations of the accelerator (we show the elements we believe are associated with

each perspective at the head of each description). The four perspectives each view

accelerators as: facilitators of new venture growth; vehicles for personal development;

communities of practice; and strategic symbols.

In outlining each of these perspectives, we also identify specific research directions

related to each that scholars might consider pursuing in order to deepen understanding of

accelerators. Note that we do not view these four perspectives as mutually exclusive (we

encourage simultaneous consideration of all these perspectives in future research), nor we do

we claim these perspectives to be the only ways accelerators can be understood (we encourage

scholars to consider what other perspectives could enrich scholarly understanding of

accelerators).

Accelerators as Facilitators of New Venture Growth

This lens views accelerators as having the following elements: Product design; Business

model refinement; Market confirmation; Non-financial resources; Exposure; Funding

opportunities; Limited duration; Networking

In light of our review of academic definitions (see Part #1 findings), certainly the most

prominent way in which accelerators tend to be presently (though often implicitly) understood

by scholars is as facilitators of new venture growth (e.g., Cohen, 2013; Cohen & Hochberg,

2014; Hallen et al., in press; Hochberg, 2016). Consequently, the key performance outcome

of an accelerator from this point of view should be the development of successful new

ventures. Interestingly, though, explicit academic definitions of the accelerator rarely make

reference to this implied purpose, tending instead to focus on the more specific aspects of

accelerators’ structure (e.g., cohort-based, time-bound), and the precise ways in which

accelerators promote new venture growth (e.g., provision of training, mentoring, and funding

opportunities). Of course, some might justifiably argue that simply defining accelerators as

“facilitators of new venture growth” is much too broad, and that further elements need to be

included to distinguish accelerators from other programmes or organisations that share this

purpose (e.g., incubators, angel investors, entrepreneurial mentoring programmes). This is a

valid point, and as mentioned at the outset of this paper, we not dispute the argument that

clear and precise definitions are certainly important for building a body knowledge on a

phenomenon (Locke, 20112; Suddaby, 2010).

Yet given the fact that the accelerator phenomenon itself continues to evolve, and the

vast range of things accelerators claim to offer ventures and entrepreneurs (see Part #2

findings), we suspect that at this stage, there may be more theoretical utility in adopting a

broader (rather than narrower) conceptualization of the function(s) of accelerators. When

framed broadly as facilitators of new venture growth, scholars are encouraged to study

accelerators in the widest sense possible. Conversely, by committing to a highly specific

definition of the accelerator, we risk producing a body of theory relevant only to a particular

form of accelerator, which fails to capture the nature and consequences of the many other

forms. For instance, 51% of the accelerator websites we surveyed claimed to offer ventures

exposure to customers, yet this feature was not apparent in any of the academic definitions we

collected, which tend towards specificity rather than generality. If a researcher were to adopt

any existing academic definition to guide selection of which accelerator(s) to study, a

substantial portion of potentially relevant cases would be automatically ruled out and

subsequently unaccounted for in the eventual theory they generate.

It is also interesting to note that the vast majority of existing research on accelerators

focuses on those that aim to cultivate growth of for-profit ventures. Indeed when discussing

questions of the value of accelerators, large technology companies such as AirBnB and

Dropbox are frequently noted as accelerator “success stories”. With that said, and while still a

fledgling area of research, scholars have in recent times begun to study accelerators with

alternative missions, including social accelerators focused on cultivating not-for-profit

ventures and social enterprises (e.g., Casanovas & Bruno, 2013; Pandey, Lall, Pandey, &

Ahlawat, 2017). Nevertheless, in light of the various “grand challenges” presently facing

humanity (George, Howard-Grenville, Joshi, & Tihanyi, 2016), it seems worthwhile to

conduct more research on accelerators that aim to grow ventures with social and/or cultural

objectives. Indeed recent times have seen the emergence of accelerators that do not even seek

to recruit and grow “ventures” in a traditional sense, but rather projects or initiatives that have

an application within particular organizations or industries. Lightning Lab GovTech, for

instance, is a prominent up-and-coming accelerator in New Zealand that was part of our

sample in Part #2, but does not admit ventures: rather it assembles teams of employees from

various government departments who then attempt to develop “innovation projects” aimed at

solving problems with a governmental/social focus (e.g., domestic violence, data privacy).

We suspect that non-traditional accelerators like GovTech will become more common in the

coming years. Thus, scholars would do well to be open-minded about what they consider to

be accelerators in the first instance, and in-turn, the types of accelerators they study.

Accelerators as Vehicles for Personal Development

This lens views accelerators as having the following elements: Moral support; Non-financial

resources; Education; Mentorship; Customizability; Application process; Cohort;

Concluding event

A second way of understanding accelerators is as vehicles for personal development of

entrepreneurs. Consequently, the key performance outcome of an accelerator from this point

of view is the growth of individual participants (Fischer-Smith, 2019). We use the term

“personal development” in its broadest sense here to refer to enhancement of entrepreneurs’

skills, knowledge, abilities, identities, and general well-being. From this standpoint,

accelerators are seen to exist less to develop ventures, and more to develop the individuals

responsible for launching and operating such ventures. Relative to the venture growth

perspective outlined above, the personal development perspective is significantly less

represented in the emerging accelerator literature. While learning is frequently discussed as a

key outcome of accelerators (e.g., Cohen et al., in press; Hallen et al., in press), such

discussions are typically in reference to learning at the venture-level rather than individual-

level, and are (arguably) more aligned with the venture growth perspective described above.

We believe there is scope for significantly more work adopting a personal

development perspective on accelerators. Indeed, from an economic perspective, aiming to

develop entrepreneurs (rather than ventures directly) seems a logical approach. Most of the

entrepreneurs who participate in accelerators tend to be nascent entrepreneurs with limited

experience launching ventures. In light of the wealth of research that attests to the importance

of deliberate practice in developing any given skillset (Ericsson, 2018), it is presumably rare

for entrepreneurs to experience immediate success with their early ventures. This is not to say

that they are and always will be poor entrepreneurs, or even that their venture was a “bad

idea”, but simply that at that particular moment in time, they lacked the requisite skills,

knowledge, and abilities to ensure the success of their venture. By focusing on developing

individual entrepreneurs (and less on the specific venture they bring to the accelerator),

accelerators cultivate the seeds of long-term entrepreneurial activity that might otherwise fail

to grow after an initial failure. Some intriguing questions scholars could consider in this

respect include: what (if any) aspects of entrepreneurs’ psychology can accelerators help to

develop (e.g., entrepreneurial passion; Murnieks, Mosakowski, & Cardon, 2012,

entrepreneurial intentions; Fayolle & Liñán, 2014), and under what conditions? Similarly,

what (if any) skills, knowledge, and abilities can accelerators help entrepreneurs to develop

(e.g., information collection, idea communication), and under what conditions? Finally, at a

more general level, what are the implications of accelerator participation for entrepreneurs’

identities, self-esteem, and subjective well-being (Diener, 2000)?

Accelerators might also help individuals develop by leading them to realise (in a

relatively low-risk environment) that entrepreneurship is not in fact a path they should or even

want to pursue long-term. In today’s cultural climate, “entrepreneur” and “innovator”

identities are associated with a certain degree of social status, which presumably motivates

some individuals to pursue such identities despite lacking the requisite skills, dispositions, or

motivation (Murnieks, Klotz, & Shepherd, 2019) for the underlying activities. In the pursuit

of such identities, which can unfold over several years (perhaps even a lifetime), such

individuals and the ventures they launch can consume significant economic and social

resources (e.g., government or investor funding). Accelerators might help to mitigate this

inefficient use of resources by acting as “filtering devices” in entrepreneurial ecosystems,

providing a time-bound, low-stakes context within which individuals can clarify the “possible

selves” (e.g., Markus & Nurius, 1986) they are hopeful (or capable) of attaining in the future.

Acknowledging this and related features of accelerators, however, requires scholars to adopt a

broad understanding of the meaning of “personal development” (i.e., not only development of

entrepreneurial capabilities, but clarification of identities and life paths more generally).

Accelerators as Communities of Practice

This lens views accelerators as having the following elements: Moral support; Exposure;

Physical space; Mentorship; Networking; In-person; Team-based; Concluding event; Cohort-

based

So far, we have introduced two perspectives on the accelerator where the primary unit of

analysis are the venture and the individual, respectively. Here we introduce a third perspective

that takes the collective of individuals who comprise accelerators as the focal unit of analysis,

and therefore promotes exploration of the cultural and social dynamics at play within

accelerators. In short, this third perspective views accelerators as communities of practice

(Brown & Duguid, 1991). Communities of practice are “informal networks bound together by

shared experiences and a passion for a particular joint exercise [and] they are held together by

strong relationships and have a shared value system, memory and knowledge base” (Angwin,

Cummings, & Smith, 2011: 160). Like any other community or social group, communities of

practice have their own cultural values, norms, and rituals that members construct, enact, and

negotiate. As a result, the key performance outcome of an accelerator from this point of view

should be related to the growth of the particular community or group the accelerator is

designed to serve, or a particular characteristic or characteristics of the group (e.g.,

cohesiveness, shared values, or identification by individuals with the group).

At present, this community-based perspective is largely non-existent in academic

treatments of accelerators, but interestingly, emerged as a feature of accelerators in our review

of practitioner perspectives, with 17% of accelerator websites including at least one claim to

offer what we termed “moral support” to entrepreneurs. This finding suggests that for those

“on the ground”, accelerators are helpful in providing a source of belonging and connection –

all factors which in-turn likely have important implications for entrepreneurs’ personal

development, and consequently, the quality of their ventures. In this way, the community-

based perspective on accelerators is very much compatible with both the venture growth and

personal development perspectives described earlier.

Given the absence of research on the sociocultural dimension of accelerators, there is

much scope for future studies of this issue. A crucial first step would be undertaking

exploratory studies that provide candid insights into the nature of social life within

accelerators — both for participating entrepreneurs, and those responsible for operating

accelerators. Qualitative methodologies, and in particular ethnography, are likely to be useful

approaches for achieving such ends given the emphasis of such methods on “thick

description” of phenomena (Charmaz, 2014). Specific questions researchers could pursue

from a community-based perspective include: what culture(s) tend to emerge within

accelerators, and how do such processes unfold? What are the (positive and negative)

implications of different accelerator cultures for the development of ventures and

entrepreneurs? How do different accelerators develop certain “organizational identities” (e.g.,

Albert, Ashforth, & Dutton, 2000), and with what consequences for the ventures and

individuals that participate in them? As an intriguing aside, pursuing these kinds of questions

could also yield valuable insights into the development of organizational cultures and

identities generally due to the typically cohort-based approach of accelerators: because

accelerators regularly “shed” most of the individuals that comprise the organization (in the

form of participating entrepreneurs), they constitute ideal settings for understanding how

organizational cultures and identities are sustained in organizations with frequent, large-scale

turnover of members (e.g., volunteer-reliant organizations).

Accelerators as Strategic Symbols

This lens views accelerators as having the following elements: Exposure; Physical space;

Education; Networking; In-person; Team-based; Concluding event; Application process

The final perspective we outline here focuses less on what accelerators achieve in a

substantive sense, and more on what social, cultural, and even political value accelerators’

existence can serve for the various parties associated with them. In this way, we are proposing

a view of accelerators as symbols that are strategically deployed both to and by various

stakeholders. Relative to the other three perspectives outlined to this point, the notion that

accelerators can have strategic value for an organization, participant, or sponsor is certainly

the most nascent, and we are yet to find any substantive treatments of this perspective in the

academic literature. A notable exception, though, can be found in a recent article by Hallen et

al. (in press), in which the authors note the potential for accelerators to serve a signalling

function for ventures, whereby mere participation in an accelerator can be interpreted by

others (e.g., investors and funders, potential customers) as indicative of the venture’s promise.

We would add that in addition to serving a symbolic function for participating

ventures, accelerators can also serve such a function for a host of other individuals (e.g., those

who found and operate accelerators), groups (e.g., entrepreneurial communities), and

organizations (e.g., companies and universities that sponsor accelerators). For example, with

an increasing number of corporate, “in-house”, and university accelerators being established

(Richter et al., 2018), a view has emerged that a primary reason such accelerators exist is as a

form of marketing — a tangible way for an organization to signal its strategic intent, and

commitment to ideals of entrepreneurship, innovation, and creativity. One can see how such a

perspective makes sense, particularly after assessing the promotional materials of many

accelerators as we have done in the present research. Accelerators often seem to be “sold” to

entrepreneurs on the basis that participation constitutes a concrete signal of entrepreneurial

intentions (i.e., that entrepreneurs are serious about pursuing their ideas and taking them to

“the next level”), and that the exposure, education and networking arising from participation

will mark entrepreneurs out as “the real deal” relative to those who are filtered out by the

application process. Relatedly, participation in the increasing number of accelerators run

within corporations and governments is often promoted to employees as a way for them to

symbolize their personal commitment to innovation and dynamism — a quality increasingly

central to the “ideal worker” moulds (Reid, 2015) that many organizations promote.

While the symbolic lens on accelerators is not one that has not been explored

substantively in the academic literature, we are confident that future research from this

standpoint would yield insights into accelerators that are both novel and enlightening. As a

starting point, we would encourage scholars to consider research questions that are perhaps

more critical of the deeper motives of those who establish, fund, manage, and indeed

participate in accelerators. In exploring individuals’ motives, scholars are likely to get closer

to uncovering the symbolic functions that accelerators might serve, which are far less

explicitly apparent (though still hinted at) in the carefully manicured public information that

is provided by and about accelerators. Successfully conducting such research, however, is

likely to be methodologically challenging, because it requires researchers to think creatively

about how to access data beyond the default rhetoric or narratives that individuals deploy

when questioned about their motives for involvement with accelerators (e.g., to learn new

skills, or promote innovation within their company), and to surface the more authentic — but

perhaps less socially desirable reasons — why accelerators exist and endure (e.g., to be

admired [or envied] in the eyes of colleagues [or competitors]). Beyond this starting point,

future research could also examine how, if at all, the images and reputations of organizations

and individuals change as a result of establishing and/or supporting accelerators.

CONCLUSION

We conclude by describing a case that highlights the significance of our central argument. In

turning the pages of a business magazine, one will likely come across Elon Musk with a

reference to the revolutionary work being done on the development of Tesla’s electric

vehicles. It is largely forgotten today, but the electric vehicle concept is not as new as many

believe. In the first decade of the 20th century, The Electronic Vehicle Company was

America’s largest vehicle manufacturer, and owner and operator of motor vehicles. In the

early years of motorized transport, electric vehicles and steam-driven cars co-existed and

competed with those driven by the internal combustion engine — they were each understood

to be viable modes of automobile. However, as David Kirsch (2000) documents in The

Electric Vehicle and the Burden of History, it was not the lack of performance that, over time,

ruled out alternative, non-combustible definitions of what a motor vehicle could be, but a lack

of influential networks, cultural values, and organizational structures. This is what resulted in

public perception converging on the combustion engine as the “one best way”. By 1912, just a

few short years after its peak, The Electronic Vehicle Company was no more.

The Electronic Vehicle Company case illustrates what we are cautioning against in

this article with regard to convergence on a single definition of accelerator. By quickly

defining a phenomenon in one way, we miss out on what could be learned by maintaining a

pluralist perspective, and observing how the phenomenon evolves in practice. With a longer

runway of open-mindedness on behalf of key decision-makers and the general public, electric

vehicles, or perhaps some form of hybrid, may have emerged as a viable alternative means of

transportation one hundred years earlier. Similarly, while it might be a more efficient means

of knowledge production to converge on a single definition of the accelerator at an early

stage, we have argued here that viewing the accelerator through multiple lenses, at least in the

meantime, is likely to promote the discovery of a wider range of theoretical insights.

More generally, the case of how we might define accelerators provides insights into

how academic research on emergent phenomena, which is particularly common in a relatively

young field like entrepreneurship, can sometimes fall victim to a rush to unitary

conceptualizations. While in the case of electric vehicles the isomorphism toward the

combustion engine was more coercive, the case of accelerator research highlights how

normative and mimetic isomorphism in academic fields (DiMaggio & Powell, 1983) can

encourage rapid convergence on unitary conceptualizations of phenomena. Assessing patterns

(and changes) in scholarly definitions of the accelerator against practitioner perspectives, as

we have done in this paper, enables us to see how scholarly norms can lead to a singular

understanding of a phenomenon more rapidly than those in the world we observe and seek to

understand. There is perhaps even a danger that we converge on unitary definitions even as

the phenomena we seek to understand become more pluralistic and complex. Indeed this may

be one important reason why practitioners sometimes see the academics who seek to study the

reality they inhabit as “out of step” with it.

Hence, the case of accelerator research emphasizes that rapid convergence on unitary

conceptualizations can come with costs, and these costs may outweigh the efficiency gains

and professional benefits of such convergence (Eisenhardt, 1989). For this reason, we urge

our fellow scholars to keep their minds open to the objects and practices they study, both with

regard to accelerators and other forms of entrepreneurial activity, and allow a plurality of

perspectives to co-exist for a little longer.

REFERENCES

Albert, S., Ashforth, B. E., & Dutton, J. E. 2000. Organizational identity and identification:

Charting new waters and building new bridges. Academy of Management Review,

25(1): 13–17.

Angwin, D., Cummings, S., & Smith, C. 2011. The strategy pathfinder - Core concepts and

live-cases. Oxford, UK: Blackwell Publishers.

Berger, P. L., & Luckmann, T. 1966. The social construction of reality: A treatise in the

sociology of knowledge. Garden City, NY: Anchor.

Bliemel, M., Flores, R., De Klerk, S., & Miles, M. P. 2019. Accelerators as start-up

infrastructure for entrepreneurial clusters. Entrepreneurship and Regional

Development, 31(1–2): 133–149.

Brown, J. S., & Duguid, P. 1991. Organizational learning and communities-of-practice:

Toward a unified view of working, learning, and innovation. Organization

Science, 2(1): 40-57.

Burrell, G., & Morgan, G. 1979. Sociological paradigms and organisational analysis.

London: Heinemann.

Calas, M. B., & Smircich, L. 1999. Past postmodernism? Reflections and tentative directions.

Academy of Management Review, 24(4): 649–671.

Casasnovas, G., & Bruno, A. 2013. Scaling social ventures: An exploratory study of social

incubators and accelerators. Journal of Management for Global Sustainability, 1(2):

173–197.

Charmaz, K. 2014. Constructing Grounded Theory (2nd ed.). Thousand Oaks, CA: Sage.

Clifford, C. (2016). Within 10 years, the number of accelerator programs in the U.S. has

increased tenfold. Retrieved from https://www.entrepreneur.com/article/271000

Cohen, S. L. 2013. What do Accelerators do? Insights from incubators and angels.

Technovation, 8(3): 19–25.

Cohen, S., Fehder, D. C., Hochberg, Y. V., & Murray, F. 2019. The design of startup

accelerators. Research Policy, 48(7): 1781–1797.

Cohen, S. L., & Hochberg, Y. V. 2014. Accelerating startups: The seed accelerator

phenomenon. Social Science Research Network.

Cohen, S. L., Bingham, C. B., & Hallen, B. L. In press. The role of accelerator designs in

mitigating bounded rationality in new ventures. Administrative Science Quarterly.

Diener, E. 2000. Subjective well-being: The science of happiness and a proposal for a

national index. American Psychologist, 55(1): 34–43.

DiMaggio, P. J., & Powell, W. W. 1983. The iron cage revisited: Institutional isomorphism

and collective rationality in organizational fields. American Sociological Review, 48(2):

147–160.

Eisenhardt, K. M. 1989. Building theories from case study research. Academy of

Management Review, 14(4): 532-550.

Ericsson, K. A. 2018. An Introduction to The Cambridge Handbook of Expertise and Expert

Performance: Its Development, Organization, and Content. In K. A. Ericsson, R. R.

Hoffman, A. Kozbelt, & A. M. Williams (Eds.), The Cambridge handbook of expertise

and expert performance (2nd ed.): 3–20. Cambridge: Cambridge University Press.

Fayolle, A., & Liñán, F. 2014. The future of research on entrepreneurial intentions. Journal

of Business Research, 67(5): 663–666.

Fischer-Smith, R. 2019. New-venture accelerators — growing human capital. Retrieved

from: https://www.digital.govt.nz/blog/accelerating-human-capital-development/

Foucault, M. 1970. The order of things: An archaeology of the human sciences. London:

Tavistock.

George, G., Howard-Grenville, J., Joshi, A., & Tihanyi, L. 2016. Understanding and tackling

societal grand challenges through management research. Academy of Management

Journal, 59(6): 1880–1895.

Gioia, D. A., & Pitre, E. 1990. Multiparadigm perspectives on theory building. Academy of

Management Review, 15(4): 584–602.

Hallen, B. L., Cohen, S. L., & Bingham, C. B. In press. Do accelerators work? If so, how?

Organization Science.

Hochberg, Y. V. 2016. Accelerating entrepreneurs and ecosystems: The seed accelerator

model. Innovation Policy and the Economy, 16(1): 25–51.

Johnson, R. E., Rosen, C. C., Chang, C.-H., Djurdjevic, E., & Taing, M. U. 2012.

Recommendations for improving the construct clarity of higher-order multidimensional

constructs. Human Resource Management Review, 22(2): 62–72.

Kirsch, D. A. 2000. The electric vehicle and the burden of history. Brunswick, NJ: Rutgers

University Press.

Kuhn, T. S. 1962. The structure of scientific revolutions. Chicago: University of Chicago

Press.

Law, K. S., Wong, C.-S., & Mobley, W. H. 1998. Toward a taxonomy of multidimensional

constructs. Academy of Management Review, 23(4): 741–755.

Locke, E. A. 2012. Construct validity vs. concept validity. Human Resource Management

Review, 22(2): 146–148.

Markus, H., & Nurius, P. 1986. Possible selves. American Psychologist, 41(9): 954–969.

Morgan, G. 2006. Images of organization (2nd ed.). Thousand Oaks, CA: Sage.

Murnieks, C. Y., Mosakowski, E., & Cardon, M. S. 2012. Pathways of passion: Identity

centrality, passion, and behavior among entrepreneurs. Journal of Management, 40(6):

1583–1606.

Murnieks, C. Y., Klotz, A. C., & Shepherd, D. A. 2019. Entrepreneurial motivation: A review

of the literature and an agenda for future research. Journal of Organizational Behavior,

(April): 1–29.

Örtenblad, A., Putnam, L. L., & Trehan, K. 2016. Beyond Morgan’s eight metaphors: Adding

to and developing organization theory. Human Relations, 69(4): 875–889.

Pandey, S., Lall, S., Pandey, S. K., & Ahlawat, S. 2017. The appeal of social accelerators:

What do social entrepreneurs value? Journal of Social Entrepreneurship, 8(1): 88–109.

Patton, M. Q. 2002. Qualitative Research & Evaluation Methods (3rd ed.). Thousand Oaks,

CA: Sage.

Pauwels, C., Clarysse, B., Wright, M., & Van Hove, J. 2016. Understanding a new generation

incubation model: The accelerator. Technovation, 50–51: 13–24.

Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. 2016. Recommendations for creating

better concept definitions in the organizational, behavioral, and social sciences.

Organizational Research Methods, 19(2): 159–203.

Reichertz, J. 2007. Abduction: The logic of discovery of grounded theory. In A. Bryant & K.

Charmaz (Eds.), The Sage handbook of grounded theory: 214–228. Thousand Oaks,

CA: Sage.

Reid, E. 2015. Embracing, passing, revealing, and the ideal worker image: How people

navigate expected and experienced professional identities. Organization Science, 26(4):

997–1017.

Richter, N., Jackson, P., & Schildhauer, T. 2018. Outsourcing creativity: An abductive study

of open innovation using corporate accelerators. Creativity and Innovation

Management, 27(1): 69–78.

Suddaby, R. 2010. Editor’s comments: Construct clarity in theories of management and

organization. Academy of Management Review, 35(3): 346–357.

Walshook, M. 2013. A systemic approach to accelerating entrepreneurship. Innovations,

8(3/4): 7–17.

Widjaja, A., & Tom, M. 2015. Pedal to the metal: Which accelerators are most successful?

Retrieved from: https://pitchbook.com/news/articles/pedal-to-the-metal-which-

accelerators-are-most-successful

TABLE 1

Key Elements of Academic Definitions of the Accelerator

Definitional

What is an Accelerator? The Case for a Pluralist Conceptualization

Abstract and Figures

Recommended publications

What is an Accelerator? The Case for a Pluralist Conceptualization

What is an Accelerator? The Case for a Pluralist Conceptualization

Startups in a corporate accelerator: what is satisfying, what is relevant and what can corporates im...

Startups in a corporate accelerator: what is satisfying, what is relevant and what can corporates im...