Tips for Developing Viral Facebook Apps

I just got back from the Community Next “Platform” conference. I chose to go to this one rather than the “Graphing Social Patterns” conference (can’t these guys come up with better names for these things?) because Community Next looked like it had more of the real players – developers and investors who had actual built apps.

There were no magic technical tips from the guys that had apps with millions of users. Actually, it was funny how random and contradictory they sometimes were. For example, some apps rely heavily on invitations, others on notifications. Often it is important to get multiple apps working together, yet different people like “Zombies” from those that like “Vampires.” As you would guess, the success of these huge apps is mostly a result of being simple, fun, viral, interesting, appealing to a large audience and most of all fast. Frustratingly, launching within the first week the platform was commonly mentioned as a big part of their success.

It was great to hear the back stories behind the guy’s apps. The most interesting common thread was that often times the apps were invented or heavily influence by a girlfriend (Honesty Box, Fluff Friends) – or group of friends. The guys all had a good sense of what would work socially, or just wrote a feature that their friends wanted (Free Gifts).

RockYou has app analysis down to a science. I will follow their lead and be more methodical in instrumenting my apps. Before starting to code, analyze all the referral points in your app. What will the messaging be,? Will it spread easily between friends? What will the artwork look like? Will it be irresistible to users.? You have to convince yourself (mathematically if possible) that each user that joins will bring at least one other user with them.

Instrument your code to track all the links into and out of your app. Do A/B testing on text and graphics. Start with a rollout of a couple thousand users (buy advertising or promote the new app from an existing app) and watch your metrics to choose the best wording and images.

The conference included a 24 hour contest to build an app and launch it to the most people possible (without doing any advertising or cross promotion). The organizers added the criteria that the apps had to be both international and educational. The prize ($5,000) was sponsored by a company called Sometrics who launched a new App tracking site at the conference. I wasn’t planning on entering, I had had a a long week. My main goal being there was to network and meet people – not code.
I can’t resist a competition though, and when I saw only about 3 to 5 people/teams were considering entering I really wanted to go for it. Still, I couldn’t think of an idea worth entering. Inevitably, without advertisements, the winner would be relying heavily on their friends to add and promote the app – and I didn’t want to spam all my friends unless I had a decent idea.

So, I headed out to the evening’s event at a local bar. I met John about 20 minutes after I ordered my first drink. He had seen I was interested in the contest earlier (I kept asking how many people were going to enter) – so we stood there and drank and brainstormed. By 10:30 we had the idea for “Dumb Americans.” It took us until about 3:00 AM to get it done, John found the artwork, wrote and researched the content for the app while I coded. The main problem was that I waited until the last thing to get the Sometrics code installed and I had a hard time getting it working (no doubt partially due to the fact it was 4 in the morning). By 4:30 our stats were tracking, and we started contacting our friends to get them to try it out.

Back at the conference in the morning, I started going from person to person getting them to try our app. It was a roller coaster of a day, we weren’t sure our stats were correct – and another team had an app that was doing well. In the end, we had more users and won the contest! I met more people by asking them in person to install my app than I would have had I not entered the contest. My Facebook friends around the world came through smashingly. Next week I’ll likely do a Zuckerbucks promotion for it to really get it going.