Choosing the right software platform is a critical task for developers. It can be a business life-or-death decision picking between Android or iPhone, Amazon Web Services or Google Cloud Platform, Xbox or PlayStation.
At the same time, the hunt for talented software developers is one of the most competitive challenges facing software companies. Their market success depends largely on persuading the above-mentioned developers to write for their platforms.
A new research paper suggests that the two sides can serve their own best interests by meeting with little sleep and lots of sugar at one of the richest breeding grounds for coding superstars: the hackathon.
These high-pressure weekend events of furious engineering can be an incredible environment for diffusion of knowledge, argues Andy Wu, an assistant professor in the Strategy Unit at Harvard Business School.
“Developers want to join platforms that other developers have already joined or will join. Then they can benefit from network effects,” Wu says. “Without seeing other developers at a hackathon, it would be hard to know how many developers are excited about a particular platform.”
A new research paper written by Wu and colleagues highlights the pivotal role played by such gatherings in creating winners and losers in the platform world. “Platform Diffusion at Temporary Gatherings: Social Coordination and Ecosystem Emergence,” published in Strategic Management Journal in August, was co-authored by HBS doctoral student Tommy Pan Fang and University of British Columbia business professor David Clough.
Wu’s research centers on strategies used by technology entrepreneurs to achieve scale for competitive advantage. Platforms are prime real estate for such explorations because, to be successful, they must win a critical mass of consumers and producers.
“If I was a tech company, I would sponsor every single hackathon I could find.”
The researchers used an innovative dataset to show that the social environment of hackathons can solve a number of problems for developers: helping them understand the difficulty of learning a new platform, the value of a new platform, and, perhaps most important, the popularity of a new platform.
“There’s been a massive proliferation of platforms, so developers cannot possibly know all the platforms that exist or have all the information they need to decide which would be worth developing on,” Wu says.
Benefitting from complex contagion
Hackathons permit what social scientists call complex contagion, in which exposure to a phenomenon from multiple sources helps speed adoption. Past research that looked at this effect from an economic perspective often assumed that developers know the costs and benefits of joining a particular platform, adds Clough. “We’re acknowledging that developers often don’t have good information.”
Wu became interested in hackathons as a doctoral student at the Wharton School of the University of Pennsylvania, where one of the largest hackathons in the world, PennApps, takes place. Fang was an undergrad who participated in and eventually helped to organize the hackathon, seeing firsthand how participants shared information with one another—whether it was a superstar developer showing off his skills to a rival team, or two coders shooting the breeze about a trendy new third-party tool over a game of Super Smash Bros. in the breakroom.
“In most software hackathons, the formal competition aspect is really just a façade to put everyone in a time-bound and high-energy environment. Most teams use the hackathon as a place where they can get a lot of work done and force themselves to learn something they wouldn’t otherwise have learned,” says Wu. “It’s actually a really collaborative environment.”
Hackathons benefit companies, too
Technology diffusion helps more than developers. Some tech companies, they observed, seem to benefit significantly by sponsoring coding events. For example, Twilio’s developer kit for sending text-message confirmations—used now by Lyft, Airbnb, and Instacart—took off after the company sponsored hackathons. Despite such success stories, Wu and his fellow researchers wondered if most tech companies were really taking advantage of the opportunities for visibility that hackathons have to offer.
To test the evangelizing effectiveness of hackathons, the researchers conducted a large‐scale quantitative study of 1,302 developers and 167 hackathons. They started by taking three years of data from DevPost, a public clearinghouse for information about hackathons around the world, including participants, prize winners, and sponsors. They then combined that with data from GitHub, an online service in which developers upload code for ongoing projects. The researchers were able to see which hackathons developers attended, and then identify what software platforms they subsequently adopted.
When they ran the numbers, they found the baseline likelihood for developers to adopt a new technology in a given year rose by about 20 percent after attending a hackathon sponsored by that technology. In addition, for every 10 percent increase in the number of participants using a particular platform at an event, developers were 1.2 percent more likely to adopt that platform themselves. “We were impressed by the magnitude of the effect size,” Wu says.
Wu and fellow researchers supplemented the data with qualitative interviews with developers, who spoke firsthand about the difficulties in deciding which new platforms to adopt, and told how they learned about new technologies while attending hackathons.
From that information, the researchers conclude that multiple things are going on at once to speed adoptions of new platforms. These include the prominence of sponsorships, in which tech companies hand out T-shirts and other swag with their name and logo to developers. Also important are observations of which platforms are used by successful peers, the direct transfer of knowledge between developers, and the general ability for developers to see what platforms are trending among the broader developer community. That last element is key to the complex contagion that leads to adoption, Wu says.
“The worst thing for a developer to do is join a platform, put in all the effort to learn it, and then have nobody else show up,” he says. “Developers want to identify and join fashionable platforms that other developers are joining as well.”
Some hackathon advice
Given the size of the effects they identified, Wu concludes companies could benefit from sponsoring hackathons much more than they currently are.
“The amount of money Google and other tech companies are putting into these hackathons is small relative to the benefit they get,” he says. “This is an unbelievably effective way to get developers to use your technology. If I was a tech company, I would sponsor every single hackathon I could find.”
Of course, given the limitations the coronavirus pandemic has placed on large gatherings, hackathons have necessarily been suspended over the past year, though Wu says some have successfully transferred at least some of their activities online to create individual competitions.
Whenever in-person gatherings resume, Wu has one message for developers who are struggling with the all-important decision about what platform to adopt for their projects: go.
“Software development can be an isolating activity,” he says. “There are relatively few outlets to gather with others. Hackathons are a tremendous way for developers to learn quickly, and for both developers and technology companies to reap real benefits.”
About the author
Michael Blanding is a writer based in Boston.
[Image: gorodenkoff]
Related Reading
- 6 Ways That Emerging Technology Is Disrupting Business Strategy
- How to Be a Digital Platform Leader
- Racism and Digital Design: How Online Platforms Can Thwart Discrimination