Dating Apps That Stop Low-Effort Likes Before They Happen
It means the heart button is genuinely locked, not just guilt-tripped into looking that way. On most dating apps, a like costs nothing — a flick of a thumb across a stack of faces, repeatable a hundred times in the span it takes to read one paragraph, with nothing in the interface checking whether you looked at anything past the first photo. A handful of products have started building the opposite rule directly into the mechanics: no engagement, no like.
Anketta is one of them, and the constraint is specific rather than vague. Before the heart button does anything, you have to mark at least one line in the other person's manuscript — their own free-form writing, not a caption under a photo — as something you liked. It underlines right there on the screen. Skip that step and the heart stays inert; it will not register a tap on a manuscript you haven't actually read.
That single rule reorders what a like means. It stops being a reflex and becomes evidence — proof that one specific sentence, not just a face, is why you're interested. The mechanic sits inside Anketta's broader text-first approach to matching, where writing carries the weight photos usually do, which only works if there's enough on the page worth reacting to in the first place.
Write something worth underlining.Because removing the cost of a decision doesn't make the decision better — it makes people careless with it, and there's a specific mechanism behind why. A 2020 study in Social Psychological and Personality Science tracked how daters' acceptance behavior shifted as they moved through a session of potential partners with nothing slowing them down. Acceptance likelihood dropped 27% from the first profile to the last — not because the later profiles were worse, but because continuous, frictionless access to more options produced what the researchers called a "rejection mind-set," a decline in satisfaction and a rise in pessimism that had nothing to do with who was actually on screen (Pronk & Denissen, Social Psychological and Personality Science, 2020).
That's the cost of a free like — not to the person receiving it, but to the person sending it. Nothing in a zero-friction interface asks you to slow down, so nothing stops the drift. A small barrier that forces a pause — read this, then decide — is a direct countermeasure to a documented failure mode, not an inconvenience bolted onto a system that otherwise works fine.
Because most of them were never earned in the first place, and the receiving end can usually tell. A large-scale analysis of a major U.S. dating platform found that reply rates fall off sharply once a message reaches for someone visibly more sought-after than the sender — for the highest tier, the reply rate never climbs above 21%, no matter how many messages get sent (Bruch & Newman, Science Advances, 2018).
"Men are more than twice as likely to receive a reply from women less desirable than themselves than from more desirable ones." — Elizabeth Bruch & M.E.J. Newman, Science Advances (2018) PMC mirror
Volume doesn't fix a reply-rate problem — it just produces more unanswered messages, faster. A like that references nothing specific reads as one more entry in that pile, and most people have gotten fast at spotting the pattern: no reference to anything they wrote, no reply.
In three steps, and none of them are optional:
- Open a manuscript. You read someone's free-form writing — no photo grid, no bio squeezed into three lines, just what they wrote about themselves.
- Mark at least one line as a "like." Select a phrase that actually caught your attention and mark it — it underlines. You can mark as many lines as you want, positive or negative, and nothing forces you to stop at one.
- Press the heart. Only now does it register. If the other person has done the same — underlined something of yours and pressed their own heart — the two hearts create a match.
There's no way to shortcut step two. The heart button doesn't check whether you liked the person in the abstract; it checks whether the system has a record of you underlining something specific. No underline on file, no working heart.
Write the paragraph someone can't scroll past.Yes — and this is the part most "friction is good" coverage gets backwards. Underlining and liking are not the same action, and the relationship between them only runs one direction. You can mark three lines in someone's manuscript, or ten, and never press the heart. Nothing about marking a line commits you to anything; it's closer to taking a note than casting a vote, and it quietly teaches Anketta's recommendation model what kind of writing tends to catch your attention. The only way to feel the difference is to open a manuscript and try it yourself.
What you cannot do is the reverse. A like requires at least one underline first — always, no exceptions, no "like anyway" button hiding behind a second tap. The asymmetry is deliberate: reading and reacting to text should be cheap and reversible, but declaring interest in a person should cost something real. Anketta keeps those two thresholds at different heights on purpose.
The difference isn't how big the gesture is — it's whether the app can verify what happened before it. A super-like or a double-tap on a photo is still zero-reading-required; it's just a louder version of the same free action.
| Mechanic | Typical swipe apps | Photo-and-prompt apps | Anketta |
|---|---|---|---|
| What a "like" costs | one flick, no reading required | a tap on a photo or prompt card | at least one marked line in the person's own writing |
| Can you like without reading anything | Yes | Yes — prompts are optional to read | No — the heart stays inert until an underline exists |
| Can you read without committing to anything | Not applicable — nothing to read before deciding | Yes, browsing prompts costs nothing | Yes — underlining doesn't commit you to a like |
| What the receiver can infer from a like | Almost nothing | Which photo or prompt got attention | That a specific sentence they wrote was read and mattered |
A super-like says I noticed you faster than usual. An Anketta like says I read this specific sentence, and it's why I'm here — a narrower claim, but a verifiable one.
Fewer, and that's closer to the point than a drawback. A gate that takes real seconds instead of a fraction of one will always produce fewer total likes than an interface built to make liking as fast as possible — that's arithmetic, not a design failure. The question worth asking isn't whether the count goes down. It's whether the likes that remain mean more, both to send and to receive.
A like that required you to find and mark an actual line is a like you can defend if someone asks why you sent it. It's also a like the other person can reasonably trust wasn't fired off to fifty other people in the same ten seconds — an inference, not a promise, but one that a 48-hour match window then gives both people time to actually test. Fewer, better-founded likes are a different trade than more, cheaper ones — not automatically the right trade for every person, but a coherent one, and it's the trade the interface is built around rather than one it apologizes for.
The broader shift away from pure card-decks — what some coverage has started calling the end of swiping — covers more than one flavor of "slower." A gated like is one specific implementation of that shift, distinct from apps that just remove the swipe gesture itself without changing what a tap actually requires.
What counts as a "low-effort like" on a dating app?
A like sent without reading anything specific about the person — a reflexive tap on a photo, a mass "like everyone" pass, or a super-like used as a volume tactic rather than a genuine signal. The common thread is that the sender couldn't say what, specifically, made them interested.
Does marking a line as a "like" commit me to liking the person?
No. Marking a line only tells the system you found that specific phrase interesting — it doesn't notify the other person and doesn't count as expressing interest in them. You can mark several lines across a manuscript and close it without ever pressing the heart.
Why do I need at least one underline before I can like someone on Anketta?
Because the heart button checks for it directly — there's no path to a like that skips it. It's the mechanical enforcement of "read before you react," applied at the interface level rather than left as a suggestion in the app's onboarding copy.
Does this mean I have to underline something before I can pass on a manuscript?
No — declining doesn't require anything. The underline requirement only gates the heart, because expressing interest is the action worth slowing down; passing on someone doesn't need the same friction.
Will a gated like get me fewer matches than a typical swipe app?
Likely fewer in raw count, since the gate removes the fastest, lowest-effort likes entirely. Whether that's worse depends on what you're optimizing for — total volume, or the odds that a given match is worth the conversation that follows.
Can I still send a low-effort like if I underline something trivial, like a single common word?
Technically the gate only checks that an underline exists, not how meaningful it is. In practice, underlining forces you to open and scroll through someone's actual writing to find something worth marking — which is already more reading than a zero-effort like on a photo-only app requires.
The reply-rate math and the choice-overload research point at the same failure: an interface that makes liking free eventually makes liking meaningless, for the person sending it as much as the person receiving it. Anketta's bet is that a heart button with one small, mechanical requirement in front of it is worth more than a hundred that light up for free.
Unsure about writing? Try reading first.