OKCupid: We experiment on users too

July 28, 2014, 11:55 PM UTC
Courtesy: OK Cupid

Facebook isn’t the only Internet company to experiment on its users. On Monday, the popular dating web site OkCupid revealed that it, too, had toyed with its 30 million members.

The announcement comes weeks after Facebook (FB) revealed it had adjusted the news feeds of over 600,000 users — dialing up or down the number of positive and negative posts they saw — without their explicit consent. The move was part of a psychological study to examine the potential effects of social media on human emotions. The study generated a hailstorm of criticism that Facebook was manipulating its users, even possibly harming them. (“I wonder if Facebook KILLED anyone with their emotion manipulation stunt,” Lauren Weinstein, a privacy activist, Tweeted. “At their scale and with depressed people out there, it’s possible.”)

In the case of OkCupid, the dating site not only acknowledges it’s performed tests on users like Facebook did, but it argues such tests are the norm. “Guess what, everybody: If you use the Internet, you’re the subject of hundreds of experiments at any given time, on every site,” Christian Rudder,OkCupid co-founder, wrote in a blog post. “That’s how websites work.”

In the same post, Rudder listed three of OkCupid’s “most interesting experiments.” In one, the web site celebrated the release of their blind date app last year by taking down users’ photos for seven hours. Unsurprisingly, people engaged with the site less overall, but they also interacted with one another online differently: People responded to initial messages 44% more often than normal, and they exchanged phone numbers or email addresses more quickly. Without photos to go on, users were forced to actually talk to one another.

Another experiment sought to analyze just how important a person’s looks were in users’ quests to find love. Based on data gathered by the site, users essentially equated “looks” and “personality” as being the same thing. Rudder pointed to a user’s profile photo of a scantily clad woman clinging to a block of wood — a photo that would not have seemed out of place if run in the latest Sports Illustrated Swimsuit Issue. That particularly fetching user ranked in OkCupid’s 99th percentile as far as popularity goes, despite the fact her profile contained no text whatsoever.

But perhaps the craftiest of tests – or most nefarious – involved gauging the power of suggestion, or in this case, the influence OkCupid has on users. The site took what it considered “bad matches” and recommended them as “good matches.” As it turned out, users were more likely to interact with those matches. What didn’t change however was the likelihood of users hitting it off. Even meddling with people’s online experience it seems doesn’t affect plain old compatibility.

Still, what OkCupid effectively did was lie to its users. It had promised to match them with the most compatible potential mates. And in the case of the last experiment, it did the opposite. Rudder can say what ever he likes — “that’s how websites work” — but when those experiments defy the site’s original goal and tarnish the user experience, that’s not just manipulation on a large-scale, it’s outright deception. More so even than Facebook’s study (Facebook played with the ratio of negative to positive posts users saw — it didn’t pass off said posts as anything else than what they were).

If OkCupid’s tactics reflect the web in general, the digital world should rethink business as usual.