How game theory improves dating apps

By Uri Bram

Traditional heterosexual dating apps have a fatal flaw: women get flooded with low-quality messages – at best vapid, at worst boorish – to the point where checking the inbox becomes an unappealing chore. Partly as a result, men see most of their messages ignored. Nobody is happy, but nobody can do anything about it. Well, none of the users, individually, can. But a new generation of dating apps impose limitations on daters that might liberate them.

The executives at the apps themselves tend to see the problem as one of gender dynamics; their innovations are intended to tackle the unhappy experiences that too many women report. Dawoon Kang, co-founder of Coffee Meets Bagel, says “the reason women haven’t been fully excited about using dating services is because there wasn’t one that understood how women want to date.” Sarah Mick, Chief Creative Officer at Bumble, says her app wants to end “digital cat-calling,” and to subtly give women more power in their dating interactions. In their efforts, both apps employ strategies that a game theorist would approve of.

Kang reports that American dating apps traditionally had a ratio of roughly 60% men to 40% women, “which doesn’t sound that extreme, but if you actually take into account activity level – guys are twice as active as women – the gender ratio becomes even more lopsided; in the active user base it’s more like 80:20.” This kind of skewed ratio can have huge effects on users’ incentives; as Tim Harford, an economist, has written, even a slight imbalance in a market radically shifts power away from the over-represented group, as they are forced to compete hard or remain single.

One way to view the problem is as a tragedy of the commons, where users acting in their (narrow) self-interest over-exploit a shared resource and therefore harm the common good, ultimately harming themselves. The classic example is overfishing: each individual fisherman is tempted to harvest the ocean just a little bit more, and improve his current catch, but if all the fishermen do so then the piscine population plummets and everyone suffers in the long run.

In the case of online dating, the “shared resource” is women users’ attention: if every man “overfishes” then the women’s attention (and patience) runs out, and the women abandon the app altogether. The men (let alone the women) would benefit from a collective agreement to each send fewer and higher-quality messages, but have no way to co-ordinate such an agreement. When Coffee Meets Bagel launched, one selling point was its enforcement of such a policy: users received just one match per day. (Coffee Meets Bagel recently switched to a model with more, but still limited, daily matches).

Perhaps the saddest part of online dating’s tragedy of the commons is that matches, unlike fish, are not remotely interchangeable. And yet, on many apps it’s difficult for one user to signal to another that he is deeply interested in her specifically and not merely trying his luck with everyone. In one sense, the problem is simply that sending messages is too “cheap” – it costs nothing monetarily, but also (in contrast to real-world dating) requires vanishingly little time or even emotional investment. As a result, not only are women inundated with messages, but receiving a message becomes a very weak signal of potential compatibility.

In theory, men can make a costly signal to a woman on any app by carefully reading her profile and sending a personally crafted message instead of a generic “hey.” But some apps give users more ways to send costly signals to specific matches. Coffee Meets Bagel has a Woo button, where users pay (with the in-app currency) to send an extra signal to a specific someone. Bumble allows men to “extend” one, and only one, match each day, which tells the recipient that she’s (at least somewhat) special to him.

Bumble’s unique feature is that only women can make the first move (that is, send the first message). Of course, this greatly restricts activity for the men, but the restriction breaks the great coordination problem and solves the tragedy of the commons: since women are not being inundated with messages, the men they match have a real chance of a date. Even for the men, the benefits may well be worth the price.

Bumble has several other features that strategically influence users’ behaviour in order to lead more users into real conversations. For example, after a match is made, women only have 24 hours to start chatting or else the match disappears. Any worries that responding too quickly will signal over-enthusiasm are allayed because it’s common knowledge that the app leaves no choice. Similarly, women don’t have to worry about how they’ll be perceived for initiating a conversation. “We have a phrase for these things: just blame it on Bumble,” says Mick. The apps strategically restrict choices to shift users out of a bad equilibrium – low-quality messages and low response rates – into a better one.

While the dating market will always have a heart of its own, many other markets face similar challenges in the internet age. In the online job market it’s trivially “cheap” to submit one more CV for one more role, so employers receive hundreds of unsuitable suitors for every open position. Online apartment-hunters and apartment-owners face similar levels of inundation and frustration. With certain tweaks, some of the strategies pioneered by the dating apps could be used in other markets. Where love leads the way, perhaps others will follow.

ILLUSTRATION DONOUGH O’MALLEY