How to Fix Location-Based People Discovery

Philip Cortes is co-founder of people discovery startup Meeteor. Follow him on Twitter @philipcortes.

No clear winner came out of South by Southwest’s battle of people discovery apps. Highlight seems to have received the best press, and according to Robert Scoble, about 5% of SXSW used the service. Despite this buzz, the consensus was that all of these services fell short of expectations.

Why did these apps fail? And what should we as an industry be focusing on? Here are some ideas:

1) Lack of Single-Player Mode. What value does the very first user to sign up for your service get? If you haven’t answered this question, you’re going to have a really hard time overcoming the glaring network effect problem of this space. Foursquare solved this issue through badges and mayorships. You could be the only person in Chicago using Foursquare, but you were still rewarded for your check ins. Furthermore, you communicated your victories (badges) outwardly by posting them to Twitter and your wall. You also demonstrated the use of the service every time you checked in at a restaurant in front of your friends. I can’t even count the number of conversations that have been kicked off about about Foursquare when I tried to subtly check in at brunch. Services like LinkedIn overcame this problem by being the first online repository for your online resume.

2) Not Capturing Intent. Two years ago, we launched the very first version of at Wharton. The goal of Meeteor at that time was to manufacture serendipity amongst students based on their calendar availability and social overlap. (The same type of algorithm driving connections on Glancee + Highlight today). The feedback from users was unanimous: “This is neat, but when I wake up Monday morning, I don’t have a pressing need to meet people around me randomly just because we have the same interests.” People would then try to give us a specific use case they would imagine using our service for: “Why don’t you sell this to universities to match roommates?” or “Why can’t I tell you exactly what I’m looking for?”

The social overlap between users can act as the lubricant that facilitates meeting, but it alone won’t compel two strangers to meet. For two people to connect, they need to have overlapping needs, desires, or wants. Two people looking to date or two people who can help each other are examples of mutual overlap. Another would be two people who can help each other professionally. Prompting me to talk to someone nearby because we share two friends in common and both are into kite surfing will never be a compelling enough reason for me to reach out to and connect with a complete stranger in life. We need to better understand our user’s needs, and help them fulfill those needs.

3) Transparent Privacy Settings. Early adopters may be willing to try anything, but in order to cross the chasm, People Discovery services need to do a perfect job of letting users control their privacy.

4) Pick a Niche. This is borrowing from Chris Dixon’s “bowling pin” strategy, which you can read here. Do you want to connect people as a dating service? Do you want to connect people professionally? Capturing intent helps answer this question, but it’s very difficult to gain mass adoption if you try to be all things to all people. Individuals will have a hard time figuring out what a good use case is for your service, and when there’s no clear use case, they’ll forget about it entirely.

5) Mimic Offline Behavior. What is it that people are currently doing, that you can do better? One interesting opportunity here is to help two people who are already chatting with one another, and providing them with additional topics of conversation. (Read: I’m at the conference coffee table and about to kick off a conversation with the guy next to me, dear iPhone, please help me come up with some solid discussion topics). Knowing that I have two friends in common with the person I’m “stuck” talking to would be a significant value add, and would certainly make me enjoy the conversation a lot more.

All five of these solutions don’t have to be solved perfectly in order for one app or web service to win the race. Finding a solid Single Player Mode for your service, for example, could be a good way to buy enough time to figure out what your deeper value proposition is. Some would argue that Foursquare is only now adding the value layers (Radar + Exploresquare) that will make their service everlasting.

A combination of several of the above points, however, is likely to be mandatory before any service crosses the chasm and gains mass adoption, Twitter-style.