How we measure

It starts with keyword searches against Telegram's public directory — a few hundred terms across 20 subject areas. That finds the first few hundred channels.

The rest comes from Telegram itself. Every channel has a recommendation list, and walking that graph outward finds channels no keyword would have surfaced. The index grows by following what Telegram already knows about which channels belong together.

Once a week each channel is resolved again: subscriber count, description, verification flags, and the timestamps of its recent posts. Nothing is re-fetched afterwards — the stored response is the record.

Growth is proportional and measured against the same channel at the previous reading, and every figure on the site carries the interval behind it. Posting cadence is computed from timestamps only.

We publish what Telegram shows any logged-out visitor to a public channel: name, @username, description, subscriber count, creation date, verification flags, and the recommendation list.

We do not publish message contents, member lists, or anything at all from private groups. Post timestamps are used to compute a rate and are not republished individually.

The index covers public channels above 1,000 subscribers that a keyword search or a recommendation edge reached. It is a large sample of Telegram, not a census of it, and the topics reflect where the seed terms pointed.

The cadence is weekly, held constant, which is what makes any two weeks directly comparable.