Hackathons are a hot commodity these days. It’s no wonder. These creative competitions are yielding big ideas on almost everything. Need to test a new product? Hack it. Curious about stakeholder engagement or user experience? Looking to train staff in new capabilities? Seeking new perspectives and new solutions to your most pressing challenges? You can hack all of those.
Once the province of students in college dorms, today’s hackathons have gone mainstream, with many of the world’s most valued companies (think Google, Microsoft, Amazon), government, academia, and nonprofits vying for answers.
There are so many problems to solve today. As data proliferates, we’re in an era of vast potential for data-driven discovery and innovation. The right algorithm could change everything. Hackathons are a great way to generate new ideas, to innovate in ways that weren’t possible before, and best of all, to build a community that inspires new approaches and incites the passion of participants.
So what is a hackathon? Why a hackathon? What can one do for your organization? How do you organize one? Who should participate?
First, a hackathon can be many things. We will focus primarily on data hackathons, though that’s not the only type. At the heart of each one lies a thorny problem to be solved … and an open data set teaming with possibilities. The best hackathons bring multidisciplinary teams to the mix to brainstorm ideas, to introduce new tools and methodologies, and to prototype solutions. It’s an experience that can drive to solutions relatively quickly when the purpose, the goals, and the rules are clearly stated.
What can a hackathon do for your organization? There are so many benefits to derive:
- Solve hard problems.
- Look for ways to improve.
- Showcase technical capability.
- Build a community.
- Attract and retain talent.
Though the focus and structure of a hackathon may vary, what’s consistent across all are the four Cs: collaboration, crowdsourcing, competition, and creative problem-solving. An event can last a few hours and involve small internal groups, or run for months and involve thousands of global participants. (The Data Science Bowl competition, described below, is an example of the latter.) There is no “one-size-fits-all” solution. However, there are tried-and-true practices that encourage both individual ingenuity and team performance.
- Start with a plan. Once you develop your challenge and can succinctly define the problem set, then get granular about gathering the data and figuring out which technologies and resources are at your disposal.
- Set the ground rules. With a few rules, hackathons can prime the pumps with new thinking. It’s key to establish clear judging criteria to influence the process and the outcomes.
- Find a great facilitator. You need someone who knows who’s participating, understands the nature of the challenge problem, is aware of data available, and can clearly speak to the desired outcomes. A strong facilitator will encourage a collaborative spirit even when the competition gets intense and help overcome any siloed thinking that may exist among a diverse set of participants.
- Share lessons learned. Follow-through is especially important with hackathons. At the end of your challenge, share the artifacts with participants and the stakeholders (which might be only the sponsoring organization or the general public, or something in between). Follow-through makes the event a force multiplier for moving ideas forward.
Start with a purpose, design the judging criteria to keep everyone on track, and determine—up front—how teams will present their findings at the end. With those three ingredients determined, you’ll create an environment where a diverse team can thrive.
So who should participate? One might assume only data scientists and coders belong in a data hackathon. While computer and analytics skills are an asset, take a more open-minded approach. Purposely look for problem-solvers of all types to participate, especially aimed at achieving a balance of quantitative and qualitative skillsets. The more diverse the participants, the more innovative and insightful the outcomes are likely to be. You want to bring different and even competing ideas into the mix. When you break down siloes, you invite the wisdom of the crowd to emerge. That’s the power of a great hackathon at work.
While a great challenge is often enough, an enticing prize may raise the stakes—so does a cause that others can really embrace. For example, some hackathons aim for social impact. My favorite is the annual Data Science Bowl which focuses on data for social good. While cash prizes are awarded to the teams who design the most promising solutions, it appears to be the challenge problem itself that most inspires participation rather than the cash. One year, the competition focused on new ideas to sustain ocean health. Other years have focused on health issues—like early prediction of lung cancer or early diagnosis of heart disease—that can affect huge swaths of the population.
What’s also exciting about hackathons is their ability to focus on serious and pressing problems. For the U.S. Navy, we designed HackTheMachine, where participants were challenged to penetrate the security systems of simulated Navy warships, with the goal of eventually creating products that would improve maritime cybersecurity measures. Another hackathon for not-for-profit Polaris was aimed at disrupting human trafficking networks by using data on illicit massage businesses to find and stop traffickers.
While guiding clients on how to use hackathons to meet their mission requirements, my own organization takes full advantage of hackathons for our own requirements as well. Internal hackathons are an excellent way to engage employees and even train them in new capabilities, including data science itself!
The bottom line is clear. When you’re looking for new ways to solve tough problems, a hackathon might be your answer. For tips on how to get started or resources to host a competition, check out Kaggle.com or the white paper, You Can Hack That: How to Use Hackathons to Solve Your Toughest Challenges.
Dr. Kirk Borne is a GovLoop Featured Contributor. He is the Principal Data Scientist and an Executive Advisor at management consulting firm Booz Allen Hamilton since 2015. In those roles, he focuses on applications of data science, data management, machine learning, and AI across a variety of disciplines. You can read his posts here.