Over the years I have tried out various distribution channels in order to find potential dates, be they offline (introductions from friends, parties, etc.) or online. Online, I have tried pretty much every single site there is to try: eHarmony, Meetic, Lavalife, Matchmaker, Craigslist, OkCupid, Match, Adult Friendfinder and many more, with varying success.
Some online sites have proven to be too much work to be worth it (e.g. Craigslist or Adult Friendfinder), while others just made introductions to grossly incompatible people (eHarmony). I did meet fantastic girls on Match which led to some of my longest and most successful relationships, but even there the process is overly arduous.
Somehow on a flight back from a conference with my friend John Myers we hit upon the topic of our respective frustrations with online dating and decided to break down the issues we observed:
- Attractive girls receive way too many emails, most of which are from bad matches
- Most personal essays are generic and do not capture the real depth of someone (e.g.; in New York everyone seemingly “works hard and plays hard”), and descriptions are not objective: everyone describes themselves as being smart and attractive – so the software can’t do the work of separating the wheat from the chaff because it doesn’t have enough real world data to be able to do so
- Filling profiles out takes too long
- There is no effective way to introduce friends to each other the way people have done since Jane Austen
We started wondering if there was a way to address those issues and here is what we came up with:
- Let users define what they are looking for in a match and the importance of each of the criteria (granted it assumes you know what you are looking for; later on we could always add a more sophisticated layer which learns what you unconsciously want)
- Crowd source the ratings of various users by having friends recommend/rate each other anonymously (to prevent the incentive to lie to flatter your friends) and without anyone ever knowing the ratings, which are all kept hidden inside the app so no-one can ever be upset
- Have an algorithm match people based on what each other is looking for based on their average rating from their friends (without disclosing those ratings), which should give a much more objective/accurate match
In other words, give people very few, high quality matches with the need to go to the site only if there is a match so it does not feel like work.
Given the social components, John & I decided to build a Facebook application: Find The One. It actually does everything we described above and we even spend a few thousand dollars marketing it, but it’s not taking off.
Given the current Facebook restrictions on virality, we don’t seem to be able to ignite usage (though we might have missed obvious strategies Facebook experts might be more familiar with). Maybe we limited user interactions too much and did not build enough of a gaming or fun component into the application.
In other words, this is a cry for help to all of you. We would love to get your thoughts, feedback and contribution!
At this point we are doing this out of intellectual curiosity – we would be satisfied to create something that makes a difference and have no expectation of monetizing the application.