The trigger is almost never the hour. It is a state, and the state usually arrives in the hour before you open anything. A ten second log kept for one week finds it, which is a week of work that saves you months of setting rules against the wrong thing.
The log
Every time you open a dating app, before you swipe, write five things in a note. Ten seconds, no reflection, no tidying.
- **Time.**
- **Where you are.** Bed, sofa, bus, desk, toilet at work.
- **What you were doing in the hour before.** One phrase. "Left the pub." "Finished a call with mum." "Nothing."
- **The feeling, in one word.** Bored, flat, restless, rejected, lonely, wired, unwanted, fine. If the honest word is "fine", write fine.
- **What you wanted from opening it.** Not what you expected. What you wanted in that second.
Do it for seven days without changing your behaviour. That is the hard instruction, because the moment you start logging you will want to log good numbers. Resist it. You are collecting data, not performing.
The five states most people find
Almost everyone finds their sessions cluster around a handful of states rather than a handful of hours.
**Unstructured evening.** No plan, nothing on, three hours between dinner and sleep with nothing shaped in them. This is the most common one by a distance.
**After something social that went fine.** You had a good night, you came home, and the flat was quiet in a way it was not quiet before you went out. The contrast does it, not the loneliness.
**After a rejection anywhere.** Not romantic rejection. A job, a friend who did not reply, a meeting where nobody backed you up. The apps offer a fast, cheap route to being chosen by somebody, and that offer lands hardest an hour after being unchosen elsewhere.
**A specific texture of boredom.** Not tiredness, not restlessness. The flat, slightly itchy state where nothing is quite worth doing. This one is often the hardest to name and the most reliable to spot in a log.
**Avoiding sleep.** Opening the app in bed is often not about the app. It is about not yet having to be alone with the day.
Late, tired and alone amplifies all five. That combination is the reason most compulsive swiping happens after 10pm, and [HALT is the shorthand for it](/spaces/halt-method-recovery).
Reading your own log
Three rules make the difference between a log and a pile of notes.
**Look at the hour before, not the moment.** The moment tells you what your thumb did. The hour before tells you what happened to you.
**Find the repeating word.** Most people write the same feeling word five or six times in a week without noticing. That word is the answer, and it is usually not the word you would have guessed.
**Sort by length, not frequency.** Your longest sessions and your most frequent sessions often have different triggers. The two minute checks are usually habit. The ninety minute sessions are the ones with a state behind them, and those are the ones worth designing against.
You may also find something you did not want to write down, which is that the trigger is a specific person, or the anniversary of something, or simply being reminded that you are single by a photograph. That counts. Write it.
Three things the log will not tell you
Worth knowing so you do not over-read it.
**It will not tell you why the state exists.** Knowing that you open the apps after calls with your mother is useful. Why those calls leave you flat is a bigger question and a log will not answer it.
**It will not separate cause from company.** Some triggers are correlated rather than causal. If you always swipe on the bus, the bus is probably not doing anything to you, it is simply where you have twenty spare minutes.
**It will not tell you what to do.** That is the next section, and it is a separate decision. Plenty of people finish a week of logging, feel usefully informed, and change nothing, because insight is satisfying enough on its own to feel like progress. Insight is not progress. Insight is what tells you where to put the effort.
One more thing. If the log becomes another thing you are failing at, stop it and do something else. It is a tool for one or two weeks, not a practice.
What to do with the answer
One substitution per trigger, decided in advance, physical, and cheap.
- **Unstructured evening:** the evening needs a shape, not a hobby. One fixed thing per week at a fixed time. [Two replacements that hold](/guides/dating_apps/deleting-the-apps-and-what-to-do-instead) covers how to pick.
- **Coming home from something good:** decide the next twenty minutes before you go out. Shower, podcast, bed. Do not arrive home with a blank.
- **Rejection anywhere:** message one existing friend about the actual thing. The urge is to be chosen, and being heard is nearer to the real need than a match is.
- **The itchy boredom:** something physical for ten minutes. Walk, washing up, press ups. It does not need to be improving.
- **Avoiding sleep:** phone out of the bedroom. This one is structural rather than psychological and it is the easiest of the five to fix.
Be honest about what this is worth. No trial has tested trigger logging for dating apps, or for much else in this area. What it does reliably is tell you what to aim at, which beats aiming at everything.
Why this matters more than a time limit
A time limit assumes the problem is duration. If your log says the sessions start after rejection, a twenty minute cap will simply produce twenty focused minutes of the same thing on the same evenings, and you will feel the same at the end of them.
It also changes what you do with a bad week. If the sessions are being driven by a state rather than a habit, then blowing through a limit is not a failure of character, it is the limit being aimed at the wrong thing, and [the numbers audit](/guides/dating_apps/are-dating-apps-making-things-worse-for-you) will tell you more than another attempt at twenty minutes.
There is a reason to prefer your own log over anything published. A 2025 review of 32 quantitative studies found only three used randomised designs, and that ["the most common scales include life problems within their measurement of problematic dating app use"](https://pubmed.ncbi.nlm.nih.gov/40608479/). In other words the standard instruments partly assume their conclusion. Your log does not. It records what you did and how you felt, and that is a cleaner measurement of you than any published questionnaire currently offers.
Common questions
**How long do I need to log for?** Seven days catches a full weekly cycle, including the Sunday, which is usually the important one. Fourteen is better if you can hold it.
**What if I forget to log until afterwards?** Log it afterwards. Late data is worth more than no data. Just mark it as reconstructed so you know which entries to trust less.
**My log says I open them at random.** Almost nobody is random. Look at the hour before rather than the moment, and look at your longest sessions rather than all of them. If it still looks random after two weeks, the trigger may be a low mood that runs all day, which is worth a GP appointment rather than a better log.
**Should I stop using them while I log?** No. A log of behaviour you have suppressed measures the suppression. Change things afterwards.
**Is this the same as knowing my triggers?** It is the only version of that phrase that means anything. "Know your triggers" as advice is useless without the mechanism for finding out what they are.
Sources
- [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, and scales that fold life problems into the measure.
- [Dating app use, psychological health and well-being: meta-analysis, Computers in Human Behavior](https://asu.elsevierpure.com/en/publications/dating-app-use-psychological-health-and-psychological-well-being-/). 23 studies, 26,068 people, worse reported psychological health among users than non-users.
- [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 network level as the strongest causal finding available in this area.
