Almost every directory displays a member count. Almost none display anything that tells you whether anyone is there. The gap between "how many joined" and "how many are active" is where most bad recommendations come from.
| Signal | Value |
|---|---|
| Timestamp of most recent post | High — the only direct evidence |
| Posts per period | High, when obtainable |
| Member count trend | Medium — growth can be inert |
| Absolute member count | Low — a floor, not a measure |
It is cumulative and never decreases, so it records history rather than state. A community that was popular in the past keeps a large number forever. Worse, it is the easiest metric to inflate, which means it is the one most likely to be gamed by exactly the listings you would rather filter out.
For most platforms, recency is available if you fetch the entity; volume usually is not, or is expensive. That asymmetry is why recency is the workhorse signal: it is cheap, hard to fake, and directly answers the question a reader has.
If you can only carry one activity number, carry the most recent one you measured and say when you measured it. An undated count is worse than no count, because it implies currency it does not have.
VIEW lists Telegram channels, groups and bots with a member-count snapshot per entry and separates entity types — which matters here, because silence means something different for a broadcast channel than for a discussion group.
Two caveats stated plainly: the catalogue is broad and includes a large adult-content category, so it is not appropriate for every reader; and no directory of this kind is complete.