I looked at this from a few angles with a couple of accounts, and a few things became obvious quickly. The thing nobody mentions is that moderation quality becomes obvious only after a month. The effect stayed again and again in a way that was hard to argue with.

I tested a handful of these back to back before deciding, which made the differences much easier to see. Consistently, the ratio of active to dormant profiles moves the numbers more than anything else. That difference held up every time even after I changed my photos.

I tracked this rather than trusting impressions here, and stopped assuming the obvious answer was right. What I found is that the share of accounts that are dormant sets the ceiling on what you can achieve. It became obvious consistently enough to trust regardless of which city I set.

Where I would focus your attention:

  • Check whether the app shows when a profile was last online — the difference between options is large here.
  • How obvious it is when someone has stopped using the app — I would treat this as a dealbreaker.
  • Worth confirming how many prompts the profile format offers — the answer varies more than you would think.
  • How the app treats people who travel — surprisingly hard to find out in advance.
  • Look at whether you can filter out inactive accounts — it is easy to test and rarely advertised.

I tracked this rather than trusting impressions in my own time, and a few things became obvious quickly. The recurring theme was how easy it is to delete an account — it moves the numbers more than anything else. It repeated fairly clearly even after I changed my photos.