A rule survives a lonely Sunday only if three things are true of it. It is written down. It has a number in it. And it does not require you to decide anything in the moment, because the moment is exactly when you are worst at deciding. Everything below is built to that specification.
First, say what you are going back for
Write one sentence before you reinstall anything, because the rules follow the goal and a vague goal produces vague rules.
"I want a relationship." "I want to go on dates." "I want to feel like something is happening." All three are legitimate and they produce completely different rulebooks. Someone who wants dates should be measured on dates. Someone who wants to feel like something is happening has just identified the thing the app was actually providing, and should be honest that this can be provided more cheaply and with less damage elsewhere.
If the sentence you write is closer to "I don't want to be the person who quit", that is not a goal. That is pride, and pride is not worth a monthly subscription.
The six rules
**One app.** Not three, not "just this one as well". Multiple apps multiply the notifications and split the effort, and the second app is almost always the one that gets opened when the first one is quiet.
**A fixed slot.** Twenty minutes, one time of day, the same time. Not in bed. Set a timer that goes off out loud in the room, because the phone's own timer is the phone talking to itself.
**A like cap.** Fifteen a session, then stop, and spend anything left on conversations you already have. Most people find matches were never the bottleneck.
**Move it, or let it go.** Decide the number now: after eight or ten messages either suggest meeting or stop replying. The endless pleasant chat that goes nowhere is where the hours actually disappear.
**Notifications off.** Entirely, not important-only. The badge is what starts the sessions you did not plan.
**A review date.** One specific date each month in the calendar. On that day, and no other, you decide whether this is still worth doing.
Write all six in a note that opens when you unlock the phone. A rule you can read at 11pm beats a resolution you made in a good mood.
Which app, if it is one app
Since the first rule is one app, the choice matters more than it would otherwise, and it is worth making it on structure rather than on reputation.
Three questions do most of the work. **Does the format have a natural stopping point,** or is it an endless deck? A daily limit built into the product does more for you than any rule you impose on yourself. **How much does it cost to say no to somebody?** Formats where declining requires nothing produce the highest volume of silence, and silence is the part that hurts. **Does it require writing?** Anything that makes a first message harder to send slows the whole thing down, which in this context is a feature.
The one criterion not to use is which app your friends say is best for meeting people. That is a claim about outcomes that nobody has measured, and it is usually a report of one anecdote.
If you are choosing between apps that all feel the same, choose the one you have never had a bad run on. Going back to the specific product you deleted at 1am carries the old associations with it, and starting somewhere neutral is worth more than any feature comparison.
The rules that do not work
"Be more mindful about it." No number, no test, no way to know when you have broken it.
"Only when I feel like it." The feeling is the trigger. This rule reduces to "always", with extra steps.
"Just for fun, no expectations." Whatever brought you to this piece was not fun, and telling yourself a story about your own motives is not a boundary.
"I'll stop if it starts getting bad again." Nobody notices the exact evening things get bad again. That is what the review date is for.
Make it physical, not moral
Every rule above works better with something mechanical behind it.
Screen time limit on the one app, with a passcode somebody else types in while you look away. Notifications off at system level so it is not one toggle from being back. The icon off the home screen so you have to search for it by name, because typing four letters is a decision and tapping a thumbnail is not. Grayscale if you want the cheapest tweak with actual data behind it: in a trial it ["cut measured screen time immediately and significantly"](https://pubmed.ncbi.nlm.nih.gov/36577008/), though it produced "no immediate causal effect of reducing usage on subjective well-being". Self-set time limits worked in the same trial, but more slowly and less well than the environmental change.
That distinction is the whole design principle. Change the environment and the behaviour follows. Change your mind and you have to change it again tomorrow.
The monthly review
On the review date, rerun the four numbers from [the swiping audit](/guides/dating_apps/how-to-stop-compulsively-swiping): hours, real conversations, dates, dates you were glad you went on. Then divide the hours by the last number, which gives you the current price of one good evening.
Compare it with last month. That is the entire review. Not how you feel about it, not whether this is the month it turns around. The number.
Signs you should come off again
Three, and they are specific on purpose.
You have blown through the slot more than twice in a fortnight. You have started opening it in a particular mood rather than at the fixed time, which means the trigger has moved and [it is worth a week of logging](/guides/dating_apps/when-do-you-open-them). Or the review numbers have got worse two months running.
Any one of those is a result rather than a failure. It tells you moderation is not currently available to you, which is useful information and not a character reference. [Deleting properly, and replacing what the app was doing](/guides/dating_apps/deleting-the-apps-and-what-to-do-instead) is the next piece of work rather than a stricter version of the same rules.
What nobody can tell you
No trial has tested moderation rules for dating apps. Not one. The evidence for self-set limits comes from general phone use, and in that trial it was the weaker of the two interventions tested.
The wider position is honest and worth repeating: a 2025 review of 32 quantitative studies found only three used randomised designs, noted that ["the most common scales include life problems within their measurement of problematic dating app use"](https://pubmed.ncbi.nlm.nih.gov/40608479/), and concluded that the concept remains debated. So nobody can tell you what a safe amount is, and nobody is going to.
Your review numbers are the only instrument in this that is actually measuring you.
Common questions
**Can I use dating apps in moderation?** Some people clearly can. Whether you are one of them is answered by trying it with numbers attached for two months, not by deciding in advance.
**How long should I stay off before going back?** There is no evidence-based figure. What matters more than the length is whether you have replaced what the app was doing, because going back into an empty week reproduces the same use.
**Is a twenty minute limit realistic?** For plenty of people, yes, with a timer in the room and the app off the home screen. If you blow through it repeatedly, the limit is not the problem, it is the finding.
**What if I meet someone? Do I keep the rules?** Delete the app, or at least the account. Keeping it "just in case" while seeing someone is the single most common route back to the old pattern, and it is unfair to the person as well.
Sources
- [Effects of grayscale mode and self-set time limits on smartphone use](https://pubmed.ncbi.nlm.nih.gov/36577008/). Grayscale cut usage immediately; self-set limits worked more slowly and less well.
- [Problematic Online Dating: Systematic Review of Definitions, Correlates, and Study Designs, Journal of Medical Internet Research, 2025](https://pubmed.ncbi.nlm.nih.gov/40608479/). 32 studies, only 3 randomised, scales that fold life problems into the measure.
- [Blocking mobile internet on smartphones improves sustained attention, mental health, and subjective well-being, PNAS Nexus](https://pubmed.ncbi.nlm.nih.gov/39967678/). Friction at the environmental level as the strongest causal finding available.
