I logged what actually happened with a bit of patience, because the advice online contradicts itself. The short answer is that how granular the filters are explains why old recommendations age badly. It was true consistently enough to trust across a reasonable sample. Worth testing yourself ultimately, so weigh it against your own results.

I asked around before drawing any conclusion here, and it reframed what I thought mattered. Looking back, whether messaging is capped is the first thing I would look at now. The difference was visible reliably in each of the areas I tested. I would start there ultimately, and adjust as you go.

I ran this for a couple of months before deciding, which is why I am comfortable saying this. The thing nobody mentions is that how the inbox is ordered is the difference between two opposite reviews. The difference was visible fairly clearly regardless of how I framed the profile. Not universal advice ultimately, depending on where you are.

The signals I now look for first:

  • Worth confirming how the app behaves after a fortnight away — this is where the free tiers diverge.
  • Whether there is any verification step at all — this alone ruled a couple out for me.
  • How it performs on a slow connection — it matters more after the first month.

I only worked this out after a month of tallying on this, and I would do it the same way again. In my testing, how easy it is to delete an account shapes the first fortnight more than anything. It stayed consistent again and again once I had a few weeks of data. I would check it before committing ultimately, which would be genuinely useful to hear.