Company blogs are often PR regurgitating machines on a WordPress install. But OkTrends, the official blog of dating site OkCupid, does more than blandly praise the service. Instead, OkCupid’s founders have compiled data from hundreds of millions of user interactions — from profile pictures to rape fantasies — to identify the habits and tendencies of those looking for love on the web. And though it leans heavily on math, it still offers plenty to enjoy for those of us who barely passed Intro to Statistics.
I spoke with OkCupid co-founder and CEO Sam Yagan about the inspiration and marketing success of the blog.
The Bygone Bureau: OkTrends isn’t your traditional company blog. What inspired you guys to focus on mining user data?
Sam Yagan: From the very beginning when we started OkCupid, we knew that we were going to collect all kinds of interesting data on our users, and that comes in a bunch of interesting ways: profile data; various information that people are always more willing to give a dating site more than any other site; the match questions people answer — the average user answers 233 questions about their own preferences and consumption. Then we have the actual actions, what people are searching, who people are messaging. So we knew our site was going to collect all kinds of data.
So this is data you are already collecting internally for OkCupid?
That’s right. The blog is not the end of what we’re trying to create. The primary use we thought we would have for the data was for advertising. Our business model is advertising based, and having all of this deep data on our users has been really value in helping us land some advertisers.
A few years ago when we had a PR firm, we started collecting little nuggets of data and using our traditional PR firm to reach out to advertisers and get them excited. And it turns out that’s a very frustrating process — it’s very slow, and hard to get people to call you back and have those conversations. So we stopped doing that about two years ago. Then six months ago, we said, “Why don’t we just take the same kind of stuff we were doing with our PR firm and start publishing it on our blog in much more interesting, much more accessible way?” So that’s what we’ve been doing.
It’s been frustrating for us to see so-called experts on TV or in magazines or on the web saying, “Here’s how you should date or online date.” Because what do they know? Even if you’re a psychologist, or whatever your background is, you have such limited data, such limited sample sizes that it’s hard for you to make those kinds of claims.
In much the same way we think the eHarmony approach to dating is wrong because they take the centralized “we know what’s best for you approach” (our matching algorithm is decentralized, “you know what’s best for you”), in that same sort of parallel, we think that people who tell you what’s best for you from an advice perspective are inauthentic. All we do is say, “Here’s what happens on OkCupid, here’s what works, take it or leave it.”
I think you pointed this out in a blog post, but the amount of data you have is just so great. There was that Gallup poll about whether you were gonna vote for Barack Obama or not. And the sample size you had was way greater than that Gallup poll.
Literally tens of millions of interactions and actions take place on our site everyday. We know what profiles you click on, what profiles you rate, we know everything. It’s a remarkable amount of data that we have. The best part is that a survey is subject to all kind of errors. The Obama election is a good example: there was a lot of debate before the election over whether people were lying to the pollsters about their willingness to vote for an African-American man to be President. So that’s the other beauty of our data. It’s not survey, it’s actual observations of people in the real world, actually dating. It’s not “Will you date?”, it’s “Will you actually send these messages or not?”
I don’t know any other web company with this much user data that’s writing a blog and putting all of this information for just for people to consume or to enjoy. Why is that you think that Facebook isn’t more open about this kind of thing or that Twitter doesn’t report on user interactions?
The best example is why doesn’t Match.com publish this kind of blog. You could argue that someone like Facebook maybe… no, you can’t.
They all have amazing data. To some extent, Trending Topics on Twitter is kind of starting to do a little bit of that. But nobody does this kind of comprehensive stuff that we do. And especially, why aren’t other dating sites doing this?
I think it’s two things. One, it’s part of the brand and the voice. OkCupid is about empowering the user. It’s authentic to what we do. If Match started doing it, they would never want to tell you that black women have the lowest response rate. Because that’s going to get them a lot of trouble, and they’re afraid that they’re going to lose a lot of revenue for that. But the biggest reason for us is that we’re, at heart, math majors; we’re data geeks.
Because it resonates so much with how we match people up. The data story is our business. Our business is that we know more about you, and therefore we can set you up with better matches. eHarmony is the exact opposite. eHarmony says, “We’re going looking into your soul, we’re going to find you your soulmate, and they’re going to come out of a black box.” For eHarmony to publish this kind of information would be counter to their entire approach to dating. But I’m surprised Match.com isn’t doing this. I think they could do it, but I think they would have their legal team reviewing every blog post saying, “You can’t do this and you can’t that.” And again that’s part of our voice and our brand.
And eHarmony and Match.com are much larger companies, in terms of number of employees.
What kind of surprised me is when I sent an interview request to talk to the person running OkTrends, I didn’t expect it to be one of the founders writing it. It’s you and [founder] Christian Rudder?
Christian does all of the writing and works on the content vision, and I just talk about it. A huge part of the founding team has a involvement in the blog. I think it shows it’s something important to us.
How do you guys decide what topic or trend to write about?
In general, the questions just come internally. In some ways, it’s not that hard. Look at the stuff we’ve handled so far: race is obviously an interesting question, age is an interesting question, what makes a good photo is an interesting question. There is no shortage of topics when it comes to dating. Just pick up any magazine, whether it’s Cosmo or GQ, and half the articles are different articles about dating or sex or relationships. It’s well known what the questions are. Sometimes users will suggest them, sometimes we suggest them, sometimes we’re just reading a magazine or a blog and an issue about dating come up and we’re like, “Oh, we should explore that.”
There’s no shortage. We haven’t even touched on that much international stuff — how does geography matter? There’s just so much to cover.
Are you guys planning on covering gay trends on OkTrends?
Yeah, we will. To date, we haven’t done much but touch on it and link off to some specific graphs, but that’s on our list of things to take care of.
Has the blog helped attract new users to OKCupid?
Oh, a ton. It’s really hard to identify just how successful it’s been. Like, The New York Times wrote an article about OkCupid, but the hook for the story was the blog. That piece not only got us a lot of traffic, but also a CNN piece. I think the way buzz and word of mouth and PR and media all work is that they build off each other. So it’s hard to pinpoint how many people signed up for OkCupid after having read the blog, but it’s definitely the case that there’s more buzz about OkCupid then there’s ever been and it’s really hard to argue it at least large part due to the blog.