ASB Assistant Professor Melati Nungsari began her research of matching markets in an unconventional way: through signing up for dating websites. “I was on all of these websites just to see how they worked, from Star Trek Dating to Farmer’s Only to JDate. I covered all cultures, religions, and races pretty well,” she says. “Before this, no researchers had looked at online dating at all.
Ten years ago, it was nonexistent.” Her recent working paper covers the tradeoffs between quality and fit in matching markets including, but not limited to, dating websites. In it, she argues that placing a higher weight on fit over quality results in a better match for you while simultaneously benefitting everyone else on the platform. In these markets, there are two types of traits by which people evaluate matches: vertical traits (quality) and horizontal traits (fit).
Vertical traits are those that everyone evaluates similarly, such as school rankings (a higher ranking being more favorable). Horizontal traits, on the other hand, represent preferences that are valued differently by different people, such as a preference for cats over dogs. Melati’s research is also unique because it presents a more realistic version of matching markets than previously seen in the literature.
“Past papers simplified these markets, assuming that people only cared about one vertical trait. In this paper, we consider a multidimensional array of traits and how placing weight on different traits creates interesting externalities.” Some of these externalities are straightforward, such as rivalry externalities. When students place a lot of weight on a vertical trait like school rankings, there will be increased competition for a relatively fixed number of spots, forcing many to accept poorer matches as a result.
But there are also less obvious externalities such as intramatch externalities, which assume that you only choose matches based on maximizing the match value to you rather than your partner. Though you may be searching for the highest-quality partners, you are more likely to be rejected if the partner finds that you are a low-quality match. Melati identifies three ways to correct for these inefficiencies: pricing, segmentation, and curation.
Pricing works by limiting self-selection into certain markets, preventing low-quality participants from lurking indefinitely in hopes of finding a high-quality match. Per-match pricing can also be effective on some platforms, especially where there is a high level of curation. Segmentation occurs when platforms cater to a specific market with similar preferences on both sides.
When horizontal traits are important, segmentation is an effective way to enable connections between participants (think back to the Star Trek Dating platform). On the other hand, segmentation is less effective when applied to vertical traits, which gives rise to “elite” platforms such as BeautifulPeople. Why? “Because people lie,” Melati says. “They lie on applications, on dating platforms, and when selling on eBay.
For example, if you’re running a school that only caters to rich people, more people will lie about their income. But you would never lie about your preference for cats over dogs, because people don’t uniformly value cat lovers.” She also sees curation, or restricting choice, as effective in some matching markets.
For example, marriage-minded dating platforms such as Match and eHarmony don’t allow their members to see the full range of potential matches, instead offering up small “batches” of potential matches for consideration. This has three benefits. First, it allows platform designers to maximize the total value of matches in the market by presenting participants with higher fit matches they may not have otherwise considered.
Second, it prevents the “paradox of choice” faced by participants who are overwhelmed with too many options. Third, platform operators can provide curation as a value-added service, and thus charge more for it. This research has implications for more than just singles. Melati believes platform designers can learn from these findings by paying better attention to the behavior of participants.
She urges managers to study how participants weigh the tradeoffs between different matches, as each person’s actions has consequences for every other member of the platform. Without knowing customers and their preferences well, managers could be missing out on crucial value-adding traits or matching methods. Worse, if participants are repeatedly presented with poor matches or feel they are lied to by others, the credibility of the platform will erode.
But what about the participants in matching markets? Melati advises giving more consideration to horizontal traits. For example, when evaluating schools, evaluate the curriculum and teacher quality in addition to the rankings. Although horizontal traits are often more difficult to assess than vertical traits, she argues that it is worth the effort to investigate matches for fit as well as quality, because it will be easier to find an overall better match that is aligned with your preferences.
As far as dating goes, Melati hasn’t needed to put this advice to the test herself. “I was already dating my husband before I started this research,” she says. However, she recognizes the benefit of platforms as a convenient way to get better matches. “Our lives have become busier, so it’s harder to meet people. There are many people out there who genuinely want to find somebody.”