What actually works for reducing your own social media use?
Changing the environment rather than relying on intention — because the products are designed by large teams to capture attention, and willpower is not a fair contest against that. The interventions with evidence behind them are almost all structural.
What works, roughly in order of effectiveness:
Removing the app from your phone. Using a service only through a browser on a computer introduces friction at every session and removes notifications. This is consistently the single most effective change, and it does not require quitting anything.
Turning off notifications entirely, not just for some apps. Notifications are the mechanism by which an app initiates contact rather than waiting to be chosen.
Greyscale, which removes the colour cues that make interfaces compelling. Sounds trivial, and studies have found measurable reductions.
Moving apps off the home screen, so opening requires searching for it — which converts an automatic action into a deliberate one.
Keeping the phone out of the bedroom, which addresses both use and sleep.
Removing the account rather than the app where the pull is strong, since reinstalling takes seconds.
Using the following or chronological feed rather than the recommended one, which removes the infinite supply of novel content.
Setting a specific purpose before opening, and closing when it is met.
What works less well than people expect:
Screen time limits and app timers, which are easily dismissed and which people reliably override — though they are useful as information rather than as a barrier.
Deleting everything and quitting entirely, which has a high relapse rate and loses genuine value along with the harm.
Willpower and resolutions, against a product optimised by experimentation on millions of users.
What the evidence actually says about outcomes. Reduction studies find improvements in wellbeing and sleep, with effect sizes that are real and modest — the framing that social media is uniformly harmful is not supported, and neither is the claim that it makes no difference.
The useful question is not how much but what kind. Passive scrolling is associated with worse outcomes than active communication with people you know, which suggests changing use rather than only reducing it.