Vote 2020 graphic
Everything you need to know about and expect during
the most important election of our lifetimes

OKCupid: We Experiment On You, Deal With It

Illustration for article titled OKCupid: We Experiment On You, Deal With It

"Experimenting on people" makes most of us feel uncomfortable, like we're rats in a lab being prodded with weird devices and fed food that makes our organs bloated. But today, OKCupid tried to put that notion to rest—or at least, it tried to explain that services need to experiment on you in order to make their products better.

Advertisement

There's been a huge outcry recently over the way that Facebook experiments on users to learn more about how to improve its product. This is very silly, of course, because companies have been experimenting on their customers for ages. That doesn't make it right, but Facebook's News Feed experiments are hardly an isolated case, and more importantly, the experimentation isn't likely to go away.

OKCupid has always experimented on users, and even promotes this aspect of its methodology in its chemistry beaker branding. What else do you think you're doing when you're giving a person the thumbs up, if not giving OKCupid another data point? Still, people don't seem to get it—so founder Christian Rudder took to the OKTrends blog with a few examples of how OKCupid has poked and prodded its users. His reaction to the whiners is pretty unequivocal:

But guess what, everybody: if you use the Internet, you're the subject of hundreds of experiments at any given time, on every site. That's how websites work.

Advertisement

Rudder goes on to unpack several OKCupid experiments and how they gave the company insight into user behavior—insights that help the company improve its service. For example, he notes what happened when OKCupid temporarily removed pictures from online dating profiles. It turns out that people dismissed each other less often, and were willing to engage and even meet up in real life at a greater rate.

Rudder told me on the phone that many time these experiments lead to what amount to very mundane changes in the service like the color of the button or the location of text. In some cases though, OKCupid, does meddle in ways people might find more offensive, like showing you hotter people to get you more engaged. And ultimately all of the decisions "play with your emotions" in that they change how you interact with people that you might fall in love or have sex with.

But the bottom line is this: You are being experimented on in a highly anonymized way at all times on the internet. It's part of the price you pay for using largely free services that are so effective. If there were no experiments, these services wouldn't be nearly as good—and you wouldn't want to use them anyway. [OKTrends]

Top image by Nick Stango

Advertisement

Share This Story

Get our newsletter

DISCUSSION

eyebreakthings
EyeBreakThings

There's been a huge outcry recently over the way that Facebook experiments on users to learn more about how to improve its product. This is very silly, of course, because companies have been experimenting on their customers for ages.

That's NOT the argument against the facebook "study". Most of the people who orignally got upset over this understand that these types of studies are ran for internal marketing departments .

The issue is that the "study" somehow made it into a a peer-reviewed journal, and claimed to have passed IRB. The outcry is that there are research ethics involved when publishing into peer-reviewed journals. The IRB should have flagged the study as unethical, and never should have been published.

There are times when you may be able to get away with not notifying your subjects they are part of a study, but there are very few instances when this is OK, and the facebook experiment wouldn't fall into that category.

And it's not about "legality" before anyone jumps in on that. It's about research ethics. They can hide the notification in EULA all day long, but that's even less ethical in my mind (you don't trick people into a study).

EDIT: Clarity