eHarmony vs L'Amore Vince: Do Compatibility Scores Predict Chemistry?

eHarmony asks new users to complete a questionnaire built around what the company calls its 29 Dimensions of Compatibility model, a battery of items covering personality traits and emotional temperament, alongside stated values about family and ambition, first assembled by psychologist Neil Clark Warren when the site launched in 2000. Most people report the test takes between twenty and thirty minutes to finish, and eHarmony will not show a single match until it's done. That upfront time investment is the company's whole pitch: the length of the questionnaire is supposed to be proof that the matches it produces afterward mean something.

The 29 Dimensions, Measured in Half an Hour

The questionnaire covers ground that most dating apps skip entirely. It asks about conflict style and how a person handles a partner's bad mood, along with what role ambition plays in how they picture the next ten years. eHarmony's algorithm weighs the answers against its own model of which traits tend to sit well together, then produces a ranked list of matches, each with a compatibility percentage attached. The percentage is presented with confidence, and for a certain kind of user, that confidence is the entire appeal: a number does the sorting so the person doesn't have to.

What a Questionnaire Can't Capture

The trouble is that a static questionnaire, however long, is a snapshot of stated preference, not a record of how two specific people actually behave with each other. A widely cited 2012 review of online dating research, published in the journal Psychological Science in the Public Interest, examined exactly this claim and found no algorithm on the market, eHarmony's included, could be shown to reliably predict relationship success from pre-interaction traits alone. The reviewers weren't arguing that personality data is useless. Their point was that compatibility gets generated partly in the interaction itself: whether a joke lands, whether a pause gets read correctly, once two people are actually talking. None of that shows up on a form filled out alone, at a desk, months before either person has heard the other's voice. The distinction matters because most people don't think of compatibility as something built through repeated small interactions. They think of it as something a good enough questionnaire should already be able to detect in advance.

A Score Built to Be Tested, Not Trusted

Some newer apps handle the same problem differently, by treating a compatibility number as a starting hypothesis rather than a verdict. A score still gets computed from answers to personality questions, the same basic method eHarmony uses, but its job is only to decide who gets introduced, not to stand in for the relationship itself. What happens after the score is where the real test takes place: matched users are required to actually talk, in stages, before either person sees a face or trades a phone number. The score opens a door. It doesn't promise what's behind it.

Four Rounds Before a Face

The structure runs in a fixed order. Round one is text-only chat, where the initial compatibility score gets its first real test against how someone actually writes, jokes, and asks questions back. Round two moves to voice, an audio-only call that reveals tone and pacing without any visual information to distract from it. Round three finally opens video, and round four, once both people choose to continue, exchanges contact information. Either person can pass at the end of any round, and nothing forces the process forward faster than either side wants to go. The photo, when it finally arrives in round three, lands on top of two rounds of actual conversation instead of ahead of them.

Where the Two Systems Actually Diverge

eHarmony's score does its work before a match ever says a word. The moment a percentage is generated, matches appear as ranked profile cards, photo included, and nothing in the platform requires two people to spend real time with each other's voice or unscripted manner before a photo starts doing the deciding. A round-based structure inverts that order. The score still exists and still matters, but it earns its credibility gradually, through actual contact, rather than asking to be trusted on sight. That difference in order, not the sophistication of either formula, is probably the more useful thing to weigh.

The Honest Answer

Neither number, on its own, predicts chemistry, because chemistry isn't a property of a questionnaire. It's a property of two specific people in an actual exchange, and no algorithm has been shown to forecast that reliably from static answers alone. What separates the two approaches has less to do with which formula is smarter and more to do with what happens to the number afterward: whether it gets presented as a finished verdict next to a photo, or treated as an opening bid that still has to survive a real conversation. That's not a knock on either company's data science. It's a structural point about when, in the process, a number is asked to do work it was never actually equipped to do.

L'Amore Vince builds its own compatibility score from personality answers for exactly this reason, but treats the number as a hypothesis that has to hold up across a text round, a voice round, and a video round before it counts for anything at all.