Online Hackathon Tips from Realtime Messaging Startup, TRNQL
After three online challenges with ever-increasing participation on Devpost, I had a chance to talk with Lawrence and Mike from TRNQL, to learn what’s worked well for them and their tips on running a great online hackathon with Devpost.
If you haven’t heard of TRNQL yet, they’re a mobile PaaS for Android, iOS, and IoT that helps you build smarter apps. It’s pretty awesome how easy they make it to integrate “contextual awareness” from sensors, phone data, and cloud services into your app.
Q: What is TRNQL and how did you get started with online hackathons?
Lawrence: My co-founder Nazmul and I started TRNQL about a year ago. We’re a VC-backed startup and still relatively small at six people. Nazmul is particularly active in the developer community, which is how we found out about Devpost. He is an instructor for the UXD course on Udacity and a moderator for the UXD G+ community. One of those folks had participated in past online challenges on Devpost and suggested we try an online hackathon as a way for people to find out about TRNQL and use our SDK. We didn’t really know what to expect, but thought it would a good thing to try.
Q: Now that you have three online challenges under your belt, what have you learned?
Lawrence: We realized this could be a good, relatively fast way to get people playing around and experimenting with our SDK. We were able to learn more about usage, where devs were getting stuck, what they liked and didn’t like, etc, and make quick changes. We were also able to broaden our reach and gather feedback from developers in emerging markets.
Q: Was there anything about the experience that surprised you?
Lawrence: I’d say we were a bit surprised by how globally distributed our participants were. We had developers from the United States, Russia, Poland, Saudi Arabia, India, etc all trying out our SDK.
Q: What were some of the tactics you used to improve participation? Any tips for other companies?
Lawrence: We started by examining a lot of other past app challenge pages and made changes between each contest based on our experience. I’d say the changes that made the biggest impact were around the theme, prizing, and marketing.
- Theme — One of the things we learned was the importance of a really good theme and crafting a compelling challenge. You want to choose a theme that isn’t too broad, or too narrow. Since we were looking for feedback, we tailored the theme to specific things we wanted to test like new features or things we were considering for our roadmap. Our last app challenge, which was the most successful so far, was to use our new “smart people” feature (a.k.a. “who’s nearby?” and “who’s doing what nearby?”) and at least two other context signals.
- Prizing — We increased our prize amounts a little bit for each online hackathon and saw increased results. You want to make sure you have reasonable prizes for 1st, 2nd, and 3rd place. In our last online hackathon, we also increased the amount of available prizes to five winners instead of three.
- Marketing — The more promotion you can do, the better. In addition to being listed on the Devpost site and promotion in their newsletter, we run local meetups twice per month and also ran some adwords and other online advertising campaigns. That helped to grow our numbers faster.
Q: Did you do anything differently in the day-to-day management?
Mike: I was generally the person managing the day-to-day for our challenges. Communication and support were definitely important for conversions. About once or twice a week I used the “updates” feature on the Devpost platform to communicate with participants. That seemed to be a good balance to stay on their radar without being overwhelming. I would provide any new information, answer questions, give deadline reminders, and collect feedback around their experience. That helped us increase the conversion rates each time.
—
Thanks again to TRNQL for sharing their experience! What do YOU think makes a successful online hackathon? Tweet us your ideas @devposthacks
