Dating apps may be a non-clearing market: congestion, cheap signaling, and why rational behavior produces bad outcomes
I spent the last several months trying to understand why online dating appears to produce so much frustration despite giving people access to vastly more potential partners than any previous matching system.
I eventually came to think the interesting explanation isn't primarily cultural or gender-specific. It's a market-design problem.
The starting point is thickness.
Matching markets generally benefit when more participants enter because the probability of finding a compatible counterparty rises. But beyond some point thickness produces congestion: too many potential transactions, inadequate mechanisms for evaluating them, and difficulty sending credible signals through the resulting noise.
Dating apps appear to combine several features that make this unusually severe:
1. The market is heavily asymmetric.
The large heterosexual platforms have substantially more men than women. That creates scarcity on one side and congestion on the other.
The same marketplace is therefore experienced as two almost opposite products.
2. Signaling is nearly costless.
A swipe or like carries almost no cost.
When expressing interest is cheap, broadly signaling interest can become individually rational. But aggregate cheap signaling destroys information content.
The receiving side then gets more approaches but less information about which approaches represent serious intent.
3. Congestion changes selection behavior.
Experimental research on online dating has found that continued exposure to large sets of potential partners makes participants progressively more rejecting.
In randomized experiments, acceptance probability fell roughly 27% from the first potential partner shown to the last.
The options themselves weren't getting worse.
Exposure to the option set changed the decision-maker.
This is the part I find most interesting: abundance can reduce successful selection rather than improve it.
4. The scarce side adapts too.
When matches become difficult to obtain, the rational response isn't necessarily to continue evaluating every match as a potential long-term partner.
A scarce match can be reclassified into a lower-commitment interaction.
So the congested side becomes more selective while the scarce side becomes less willing to treat the matches that clear as serious candidates.
Neither side needs to be behaving irrationally or maliciously.
Each side is responding rationally to its own incentives.
Yet the aggregate market clears worse.
5. The intermediary has a peculiar objective function.
Historically, intermediaries in courtship—friends, family, community, school, church, neighborhood—had reputational exposure to the outcome.
Modern platforms largely disintermediated those institutions.
But the replacement intermediary has an unusual economic characteristic:
Its revenue is earned while the search continues.
A successful terminal match removes two customers from the market.
That doesn't require anyone inside the company to deliberately prevent successful relationships. It simply means that engagement and successful clearing point in different directions as optimization targets.
6. We therefore measure almost everything except clearing.
Dating companies can measure registrations, active users, likes, matches, conversations, retention, payers and revenue per payer with enormous precision.
What remains remarkably difficult for an outsider to determine is the obvious denominator:
What percentage of people entering the system successfully leave it because they found the durable relationship they wanted?
Hinge is the especially interesting case because the brand promise is literally Designed to Be Deleted.
Yet the public operating metrics overwhelmingly measure people remaining, returning, engaging and paying.
There is some offline feedback—Hinge's "We Met" feature can ask whether a match produced a date and whether someone wants another date—but that is very different from longitudinally measuring relationship formation, duration, permanent successful exits and reactivation after dissolution.
That brought me to a broader hypothesis:
The public "gender war" around online dating may partly be the social symptom of a market-design failure.
Two populations experience radically different sides of the same mechanism.
Both possess accurate information about their own experience.
Neither sees the system producing the other side's experience.
So each concludes that the other population is the problem.
I ended up writing a much longer piece tracing this through matching-market economics, signaling theory, behavioral psychology, the history of courtship, the disappearance of social intermediaries, and eventually the financial statements of Match Group.
The last part became a public-equity short thesis because I realized the sociology generates financial predictions.
If the underlying marketplace is structurally impaired, eventually I would expect to see:
- payer attrition;
- heavier monetization of the participants who remain;
- difficulty expanding the total category;
- growth increasingly sourced from geographic expansion rather than deeper successful adoption;
- and eventually a lower terminal valuation for the companies operating it.
That makes the public company an interesting way of putting an otherwise difficult sociological hypothesis under an empirical clock.
The full essay and sources are here:
https://dljlevfin.substack.com/p/the-undisclosed-denominator
I'm especially interested in criticism of the behavioral mechanism rather than the stock call.
Where does the causal chain break?
Is congestion actually the right framework?
Does cheap signaling necessarily degrade matching efficiency here?
And most importantly: what metric would you use to distinguish a dating marketplace that generates enormous engagement from one that actually clears successfully?
