Take a design and social entrepreneurship challenge and put it online – at the same time as it's running in person. Sounds a little crazy, no? In the spirit of prototyping, our scrappy team of edu-innovators gave it a try – and here's what I learned.

The Future of Stuff was our first experiment with an open online challenge, focused around innovation in making and manufacturing. We recruited young innovators from around the world to take a swing at the challenge, as we took on a challenge of our own: we had three months to create, design, and run this 14-day pilot. I was the only technical person on our original team of four, so I did research, content creation, information architecture, visual design, and front-end development. (The phrase "I'll sleep when it's over" ran through my head many times.) We created 2 main assets: a resource website and a discussion platform. On June 19, we hit the big red button: the challenge was on.

Since The Future of Stuff was a prototype run by a very small team, crystallizing our new knowledge was crucial – we were extremely interested in what we could learn.

Here are 4 of my top insights. I hope they're useful to anyone else planning an online learning experience:

1. Expect the lags and silences of asynchronicity

Asynchronous conversations take minutes to participate in but days to complete. Nobody checks in constantly. They have lives to lead.

How we learned this: I was getting emails late at night from a team in a different time zone. I wrote back the next morning. They wrote back the next night asking for clarification. The next morning, I wrote back. Time elapsed: 2 days. And this was for a step that we'd hoped would take less than thirty minutes to complete! It's hard to build momentum when you have to wait 8 hours for a response to an email.

2. Let the behavior you want, not the features you want, guide your tech decisions

Even if existing technologies aren't perfect, they may be good enough. Take advantage of their familiarity, their ubiquity, their stability, and their unwritten rules.

How we learned this: We designed a custom message board called Breaker Central for participants to post their work on. We were going to just use Google+, but a developer joined us at the eleventh hour and offered to build us a discussion platform. We jumped at his offer. But, when users first started using the platform, they had problems from the outset. Some couldn't figure out how to post. Some didn't know that they were "allowed" to post.

A very, very rough discussion board wireframe.

Though I made the design as conventional as I could, I didn't have time to write complete documentation for our new platform and had to answer questions on a case-by-case basis. Google+ didn't have all the features we wanted, but at least people either know how to use it or can, ah, google their questions. More importantly, people know the social ground rules for posting on Google+; at the very least, its similarity with Facebook gives a lot of clues.

3. Design 'Day Zero' not just 'Day One'

An online experience starts well before day one – partly because of the asynchronous lag, partly because people need to plan in advance to fit the course in with their lives.

How we learned this: The first problem was one of information overload. In a post-challenge user interview, one user said that getting a dump of tons of content at once didn't let her fully digest it. She would prefer to have content to chew on over a longer period. Her comments made me think of Hack Design, which delivers 1 design lesson a week to your inbox.

The second problem was a stupidly prosaic one. So I opened a gmail account just for the class. I signed up for a bunch of services – Vimeo, Google+, YouTube, you name it. Then the challenge started and I tried to email over 100 people at once. Guess what? My account was suspended. Should have seen that coming, but oh well. But, since the shutdown happened in the crucial first hours of the challenge, we lost some incoming emails. Though we apologized, it was clear to participants that we were scrambling to fix the situation – not an auspicious or trustworthy start.

4. Emphasize overall structure over content
People are curious how an online experience works, first and foremost. A confused online challenge participant is less likely to ask clarifying questions and more likely to just lose interest.

How we learned this: We expected to be fielding questions about content from participants, but most of what we got was logistical. Can I start late? Can I work with a team? What is the end product? How many hours does this take? Is credit offered? How do I sign up? Within a day of the challenge's start, I put up a new front page clearly answering all these questions.

Version 2.0 of the front page, foregrounding the course logistics.

But wait, there's more! Check out 4 more lessons I learned while running an online course.