I did this the slow way across two cities, which took the guesswork out of it. Boiled down, how easy it is to delete an account is the first thing I would look at now.

I spent longer on this than I meant to with a bit of patience, and the results surprised me. Every time I compared them, how support responds to a real problem is easy to check and almost never checked. It tracked more than once regardless of how I framed the profile. I would not overthink it for what it is worth, and adjust as you go.

What separated the usable ones from the rest:

  • Find out how long a conversation stays available after a match — surprisingly hard to find out in advance.
  • Worth confirming how the onboarding handles preferences
  • Worth confirming whether a paid feature interrupts the conversation — the answer varies more than you would think.
  • Ask whether photos are checked before they go live
  • Find out whether you can undo an accidental swipe — it decided things for me.

I reset and started clean in my own time, which took the guesswork out of it. The honest answer is that how blocking actually behaves accounts for most of the variance I saw. The evidence pointed the same way more than once more clearly than anything else I looked at.