What we learned launching Coderoulette

And why we shut it down two weeks later (Oct 11, 2016)

A few weeks ago we launched Coderoulette, an experiment on collaborative learning. How it worked was simple: we paired two developers in real-time and gave them a coding challenge to solve using video conference and a collaborative IDE.

The launch campaign was a huge success and we would like to share some metrics and highlights with you:

• We got featured in Product Hunt, The Next Web and Wired.
• More than 2,500 people read our launching article.
• Over 15,000 people visited the website.
• Almost 3,000 people signed up.
• More than 1,000 teams worked on solving a coding challenge.
• Over 50 teams completed several coding challenges.

That all happened over the span of 1 week. After that, the project died, and today we are shutting Coderoulette down. That was the intention from the beginning and we want to explain why.

We never saw Coderoulette as a product or business opportunity. We always called it “an experiment”. We have been working on collaborative learning for several years and this was our attempt to prove that online synchronous collaborative learning was not only possible, but also fun. We proved it.

Once we validated our hypothesis, we didn’t have any reason to keep Coderoulette running. Being able to match developers in real time means you need to get people to visit your website every minute. Otherwise people get tired of waiting to be assigned to a coding partner.

Also, keep supporting Coderoulette would mean that we can’t focus 100% on what we really care about — building the most scalable and affordable school using online collaboration (more on this soon).

So, what did we learn building and launching Coderoulette?

• People are much more engaged by experiences that require them to interact with other people than by things they do alone.
• People are less likely to drop-off if there is a person on the other side who depends on them.
• Matching people in real-time is a pain in the ass.
• There is a lot of potential in peer-to-peer learning.
• Some people want to learn. Some people are just curious. When you match someone highly motivated with someone that is just curious things don’t work. You need to consider expectations and implication.
• If the success of your product is conditioned to a great launching campaign, you are probably screwed. We got lucky, and we didn’t have many other ways of bringing people together to collaborate synchronously. If you optimize for the launch, make sure you have a plan for the after-launch.
• We LOVE seeing people collaborating, and we want to create collaborative experiences that bring people together. In fact, we will make that the DNA of our work for years to come.

If you want to stay updated and learn more about the online school we are building, follow me on Twitter or Medium.