A dating profile reads like a résumé for a job nobody has actually defined. Reliable, ambitious, kind, a good listener: adjectives so generic they could sit on top of several thousand other people’s profiles without anyone noticing the swap.
I do not think this is a design flaw exactly. It is closer to what the format is built to reward, and it quietly trains us to go looking for a set of qualities rather than a person, which might explain why so many people, once real love actually shows up, do not recognize it.
I have watched friends describe, almost apologetically, a person who checked none of their usual boxes and still made them feel more like themselves than anyone who had. The apology is the part that interests me. Somewhere along the way we decided that love arriving off-list needed an excuse, rather than treating it as information about how badly the list was ever going to work.
The résumé problem nobody applies for online dating
Researcher Eli Finkel and his coauthors, reviewing the psychological research on online dating, describe the mechanism plainly: “people seek partners on the basis of searchable attributes like income and religion rather than experiential attributes like sense of humor and rapport.” Searchable attributes are the ones a filter can sort by. Experiential ones are the ones you can only actually notice by being in a room with someone, which a profile, by definition, cannot offer you.
What this produces is shopping logic applied to something that was never really a category of goods. You can filter for height, education, and a handful of stated interests. You cannot filter for the particular way a person goes quiet before they say something they actually mean, which is usually the thing that ends up mattering.
More choices did not fix this, it just multiplied the problem
Finkel and his coauthors also make a point that the format’s defenders tend to skip past: having access to more potential partners does not make people better at choosing among them, it tends to make the choosing itself more effortful and more comparison-driven. Once you can see a hundred profiles in an evening, the temptation is to sort faster, not slower, which means leaning even harder on the traits that are quickest to compare. Height sorts in a second. The way someone laughs does not sort at all, so it quietly drops out of the process before it ever gets a chance to register.
I don’t think this makes anyone shallow — I think it makes the format itself a poor instrument for the actual thing it claims to help with, in roughly the way a résumé is a poor instrument for predicting whether you’d enjoy sitting next to someone at work every day. It was never built to carry that information, so it is strange that we keep expecting it to.
What actually makes someone recognizable as themselves
An interview on friendship, given by Alexander Nehamas, has stayed with me since I first read it. He’s talking about friends specifically, but the mechanism he describes translates cleanly to romantic recognition: it’s closer to recognizing a painter’s individual style than to checking off a list of virtues.
“A friend is a person who, in your individual actions, can see you as a person.”
You can tell a Picasso from a Bellini without either painting being labeled, because the recognition sits in something more specific than a checklist of technique. Nehamas’s point is that the people we actually love get recognized the same way, through a pattern too particular to reduce to a list of adjectives, and a profile built entirely out of adjectives has no way to transmit that pattern at all.
Why the checklist keeps outvoting the person standing in front of you
My own research sits closer to place attachment than to romantic attachment specifically, but the underlying question is the same one: why do certain things register as belonging to us and others do not, no matter how objectively good they are on paper. A place can meet every practical criterion and still fail to feel like anywhere in particular. I think people do this too, more than we admit. Someone can match every stated preference on a checklist we built for ourselves years ago and still not register, on contact, as anyone specific. Meanwhile someone who fits almost none of the criteria can suddenly feel unmistakably like a person rather than a category, and we are often slower to trust that second signal, precisely because it did not arrive through the filter we spent years building.
This isn’t an argument against having preferences — it’s an argument against mistaking the preferences for the actual mechanism of recognition, which appears to run on something closer to specificity than to matching criteria.
It also explains a confusion people describe often, after meeting someone who does not fit their stated type at all. The checklist has nothing to say about what just happened, because nothing on it was violated in a way the person can point to, and nothing on it was satisfied either. The feeling shows up anyway, and because it did not arrive through the expected channel, it often gets dismissed as a fluke, or as chemistry that will not last, rather than treated as the more reliable signal it might actually be.
The trait no app would ever think to list
What I keep coming back to is how small the actual signal usually is. Not a shared list of five interests. Something closer to the particular half-second someone takes before they laugh at their own joke, or the exact way they go quiet in a waiting room instead of filling it with talk. None of that is filterable, searchable, or listable as a strength, and I suspect that is exactly why it is the part that makes a person feel like themselves instead of like a well-assembled candidate.
I do not think the solution is to delete the apps or distrust every list of preferences we have ever made for ourselves. It is closer to holding the list loosely, the way you would hold a map you know was drawn before you actually visited the place. Useful for getting somewhere. Not the same thing as the place itself, and not something to argue with once you are standing in it.
If you have ever met someone who fit almost none of your stated criteria and still felt more like a person than anyone who did, that is probably not a mistake to correct. It might be the only part of the process that was actually working.