Platform diffusion at temporary gatherings: Social coordination and ecosystem emergence

• Date: 01 February 2021
• DOI: http://doi.org/10.1002/smj.3230
• Published date: 01 February 2021

RESEARCH ARTICLE

Platform diffusion at temporary gatherings:

Social coordination and ecosystem emergence

Tommy Pan Fang

1

| Andy Wu

2

| David R. Clough

3

1

Technology and Operations

Management Unit, Harvard Business

School, Harvard University, Boston,

Massachusetts

2

Strategy Unit, Harvard Business School,

Harvard University, Boston,

Massachusetts

3

Organizational Behaviour and Human

Resources Division, Entrepreneurship

and Innovation Group, Sauder School of

Business, University of British Columbia,

Vancouver, British Columbia, Canada

Correspondence

Andy Wu, Strategy Unit, Harvard

Business School, Harvard University,

15 Harvard Way, Morgan Hall

243, Boston, MA 02163-1011.

Email: awu@hbs.edu

Funding information

HBS Division of Research and Faculty

Development; Social Sciences and

Humanities Research Council of Canada

Abstract

Research Summary: Software platforms create value

by cultivating an ecosystem of complementary products

and services. Existing explanations for how a prospective

complementor chooses platforms to join assume the

complementor has rich information about the range of

available platforms. However, complementors lack this

information in many ecosystems, raising the question of

how complementors learn about platforms in the first

place. We investigate whether attending a temporary

gathering—a hackathon—impacts the platform choices

of software developers. Through a large-scale quantita-

tive study of 1,302 developers and 167 hackathons,

supported by qualitative research, we analyze the multi-

ple channels—sponsorship, social learning, knowledge

exchange, and social coordination—through which

hackathons serve as a social forum for the diffusion of

platform adoption among attendees.

Managerial Summary: A software platform such as

Windows, iOS, or Amazon Web Services relies on

third-party developers to create applications that com-

plement the platform and make it valuable for end

users. However, developers face a wide range of possi-

ble platforms, and they may have limited information

about which platforms would be worthwhile to develop

for. A software platform business can educate and

Received: 25 February 2019 Revised: 4 August 2020 Accepted: 6 August 2020 Published on: 6 October 2020

DOI: 10.1002/smj.3230

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and

reproduction in any medium, provided the original work is properly cited.

© 2020 The Authors. Strategic Management Journal published by John Wiley & Sons Ltd.

Strat Mgmt J. 2021;42:233–272. wileyonlinelibrary.com/journal/smj 233

encourage developers to adopt their platform by

supporting in-person software development competi-

tions, known as hackathons. Developers learn about

prospective platforms that advertise at the hackathon.

Developers also learn whether and how to use a plat-

form by observing and teaching one other. Hackathons

are particularly useful for spreading platform technolo-

gies: developers prefer to adopt widely used platforms,

and hackathons permit developers to identify and join

fashionable platforms.

KEYWORDS

complex contagions, hackathons, innovation ecosystems,

multisided platforms, technology diffusion

1|INTRODUCTION

Software platforms create value for their users by cultivating an ecosystem of complementary

products and services (Brandenburger & Nalebuff, 1996; Cennamo & Santalo, 2013; Jacobides,

Cennamo, & Gawer, 2018; Kapoor, 2018; Tiwana, 2015). Scholars of platform ecosystems there-

fore study how complementors select one or more platforms to join (Boudreau, 2010; Church &

Gandal, 1992; Gawer & Henderson, 2007; Zhu & Iansiti, 2012).

Studies of complementor decision-making often assume the complementor possesses

rich information on the platforms they are choosing between, for example, information

about the user base, competitive intensity, and governance rules. While that assumption

may hold in certain well-established ecosystems, in other settings complementors lack the

rich information they would need to make a calculative choice between alternative plat-

forms (Dattée, Alexy, & Autio, 2018). In emerging or fast-changing ecosystems, the set of

platform players may be ambiguous and data on platform usage may be unreliable or nonex-

istent (Dougherty & Dunne, 2011; Hannah & Eisenhardt, 2018; Wade, 1995). In the ecosys-

tem of general software programming—in which software developers are complementors—

the domain of possible platforms a developer can use is huge and the range of possible

applications they could develop is vast (Eisenmann, 2006; Kapoor & Agarwal, 2017). Thus,

platform research must address the question of how complementors learn about platforms

in the first place.

In fast-changing ecosystems, technology platform companies proactively market

themselves to prospective complementors (Cusumano & Gawer, 2002). These companies

sometimes use in-person events to attract complementors and to motivate them to build

products that complement the platform (West & Wood, 2013; Yoffie, Casadesus-Masanell, &

Mattu, 2003). Although prior research has noted the existence of these events, and flagged

their possible influence on complementor behavior (e.g., Dattée et al., 2018; Özalp,

Cennamo, & Gawer, 2018), the impact of temporarygatheringsonplatformgrowthhasnot

yet been studied in detail. In this paper, we undertake such a study. We use the term tempo-

rary gatherings to encompass formally organized in-person events lasting between several

234 FANG ET AL.

hours and several days.

1

We address the questions: How does attendance at a temporary

gathering affect the platform adoption decisions of a complementor? And can platform

owners cultivate an ecosystem of complementors by sponsoring these events?

The impact of temporary gatherings on platform adoption is theoretically interesting for two

main reasons. First, although theories of technology diffusion emphasize social interactions as

the basis of information flows, it is unclear whether transient interactions between strangers

influence their subsequent platform adoption choices. Prior work suggests strong network ties

are the most influential in the choice of a technology (Suarez, 2005; Venkatraman & Lee, 2004).

Multiple, persistent sources of influence may be needed before people change their behavior by

adopting a new technology (Centola, 2018). However, encounters at temporary gatherings tend

to be transitory, weak-tie interactions; as such, they may not leave much long-term impact on

an attendee (e.g., Ingram & Morris, 2007).

Second, if temporary gatherings do impact complementor behavior, then they may allow

technology to diffuse between the local clusters that make up a large-scale social network

(Granovetter, 1973; Gulati, Sytch, & Tatarynowicz, 2012; Watts & Strogatz, 1999). Prior research

has found that technology use tends to be homogeneous within a social network cluster, with

limited diffusion taking place between clusters (Lee, Lee, & Lee, 2006). For example, in a classic

study of contraception techniques, Rogers and Kincaid (1981) found that most women within a

given village use the same form of contraception, but between villages the method chosen var-

ied. Technology diffusion has been characterized as a “complex contagion,”defined as a behav-

ior that only spreads when prospective adopters are exposed to multiple sources of social

influence (Centola & Macy, 2007). Because digital platforms exhibit network effects (Church &

Gandal, 1992), they are particularly likely to be subject to such threshold-based adoption

dynamics (Granovetter, 1978). Current diffusion research takes a strong interest in identifying

interaction structures that spread complex contagions (Assenova, 2018; Beaman, BenYishay,

Magruder, & Mobarak, 2018). A distinctive attribute of temporary gatherings is that they bring

people from different social clusters together. If attendees do influence one another, then tem-

porary gatherings might be a key mechanism by which platform adoption spreads widely

through a population of prospective adopters.

We focus on temporary gatherings in the form of software development hackathons,

which we further describe later in the paper. In short, hackathons bring developers together

to create new software applications in a short time frame (Lifshitz-Assaf, Lebovitz, &

Zalmanson, 2020).

2

Software developers come to these events to learn new skills, create new

applications (with teammates), and compete for prizes. Importantly for platform owners,

these software developers are prospective complementors in the ecosystems of digital plat-

forms (Eisenmann, 2006). As sponsors of the hackathon, platform owners provide financial,

in-kind, and in-person logistical support to attending developers. Platform owners use these

events as an opportunity to promote their platform to third-party software developers (Parker,

van Alstyne, & Jiang, 2017).

Theoretically, a hackathon serves as a social context in which attendees observe and influ-

ence one another, creating the opportunity for attendees to form network ties with each other

1

Besides hackathons, other temporary gatherings include conferences, music festivals, art fairs, and trade shows.

2

Although hackathons originate from software development culture, they are now used for creative problem solving in

a variety of other managerial contexts, such as strategy formulation, brand transformation, and general product

development (Arena, Cross, Sims, & Uhl-Bien, 2017; Flores, Golob, Maklin, & Tucci, 2019; Frolund, Murray, &

Riedel, 2018).

FANG ET AL.235