The year was 2010.
We were searching for the next big Facebook app, high off our previous successes. This was when Facebook just released the News Feed, and everything from cute icons to animated fake pets made millions from ad impressions and virtual pet food. People were beginning to understand what “viral” meant in the face of hundreds of millions of users, most of whom encouraged each other to use simple, stupid apps that were honestly half advertising.
We were one of the few who got how Facebook worked. Now it was just a matter of leveraging our expertise on 1-to-n unsuspecting industries.
So we decided to get into online dating. Free online dating1.
Of course, online dating is a mature, well-worn market with a handful of big players (match.com, eHarmony) and a few up-and-comers (OkCupid). They weren’t that well designed or engineered, and they cost a ton just a few months of access. We could build a better product faster, cheaper, and instantly fill it with users.
So we did.
We assembled a pretty nice team to realize this vision. There was a designer who studied how other dating sites drew from the playfulness of their interface to encourage connections and chats. Our product lead came up with a questionnaire to test new users and whether we could derive a better match algorithm. I was building the front-end to be as dynamic and client-driven as I could given the browsers of the day. Our in-house SEO expert figured out ways to migrate our current Facebook userbase onto this app, and what kind of ad inventory we’d be able to sell once there were a few million MAUs2.
A few short months later, we were ready to launch.
At first, there was a trickle of users. We discovered early on that the users of our current apps didn’t convert well to this kind of dating, and the News Feed was useless as announcing that you’re using a dating site did and still does carry a social stigma. Facebook ad conversion was poor, and even with our dozens of (fake) seed profiles, there simply weren’t enough users to make matches possible.
I remember how we all got excited when a guy and a girl actually were chatting to each other via the chat interface, and we were querying for their conversation in the database for a few days3.
After a month of adding features, trying ways to grow the userbase and get more engagement from our existing users, we shut the entire thing down and wiped the codebase.
Turned out, one of our execs met someone in the online dating industry, and they gave us a piece of data we never considered: the cost to acquire a user who’d use an online dating site was pretty damn high. I forget the exact number, but it was enough where $60 for 3-months access (what match.com was charging) barely broke even, and others were losing money hand-over-fist just to buy market share4.
We might have been incrementally better, but the fundamentals of the business all but ensured failure, and we weren’t radical enough to create a new business model. Sometimes, naivety provides a new perspective and creates a new attack angle for a “solved problem.”
But most of the time, pretending everybody else isn’t as good just means you don’t know enough.
I can already hear those of you in the dating industry snicker.↩
Monthly Active Users, the only currency with clout in the Facebook world at the time.↩
In retrospect, this was a gross violation of the expectation of privacy, but we probably drafted the TOS to let us do things like that.↩
Incidentally, Facebook gaming was another market that had the same structure, and many gaming companies did not realize that Zynga was pricing their users out out reach long before they spent a boatload of money buying less users than they thought.↩