tgsonar

Telegram Reach vs Subscribers: What 20,013 Channels Show

A channel with 1,000 subscribers reaches 54% of them. A channel with a million reaches 5%. We measured both numbers across 20,013 channels and worked out what that means for ad pricing.

Ask what a Telegram channel is worth and you will be told its subscriber count. It is the wrong number, and it gets wronger as the channel gets bigger.

We measure both figures for every channel in the catalogue: how many people are subscribed, and how many views a typical post collects. The ratio between them is the share of the audience that actually reads. Here is what it looks like across 20,013 channels with measured views.

Reach collapses as channels grow

SubscribersChannelsMedian subscribersMedian viewsShare reading
under 1,0004,99037720353.8%
1k–5k6,8482,16745320.9%
5k–20k4,3489,1561,17712.9%
20k–100k2,72736,9354,19511.4%
100k–500k843162,58517,45010.7%
over 500k2571,033,07152,9505.1%

A channel with 400 subscribers shows a post to half of them. A channel with a million shows it to one in twenty. That is a tenfold difference, and it is not an anomaly — it is the rule. The bigger the channel, the larger the layer of people who subscribed once and never came back.

What this does to pricing

Stop comparing price per post. Compare cost per thousand impressions:

CPM = post price ÷ (median views ÷ 1000)

Two channels both ask $250 for a post.

  • The first has 300,000 subscribers and 30,000 median views. CPM = $8.33.
  • The second has 40,000 subscribers and 12,000 median views. CPM = $20.83.

By subscriber count the first looks seven times larger and therefore seven times better value. By reach it is two and a half times cheaper — real, but nothing like seven. And if that first channel turns out to pull 8,000 views instead of 30,000 — entirely normal for a channel that size — it becomes the more expensive of the two.

Use the median, not the average. One viral post drags the average up and describes a week that will not repeat.

Never take the number from the media kit

"Reach: 200,000" in a price list is written by the owner. You can check it yourself, free, without asking them: open t.me/s/<channel> — the public web view of the feed. A view counter sits under every post. Take the last twenty, drop the best and the worst, look at the middle.

That is the number you divide the price by.

Three things to read alongside reach

View spread. If every post lands within a few percent of every other, that is the signature of automated view inflation, not a live audience.

Posting frequency. A channel shipping a hundred posts a month buries your ad within the hour. All else equal, a quieter channel delivers better.

Date of the last post. Median views are computed from recent posts. If the channel went silent a month ago, the number describes history.

Where the normal range sits

Median engagement across our base is 19.8%. The lower quartile is 7.9%, the upper 49.2% — a quarter of channels sit below eight percent, a quarter above forty-nine.

There is no single norm, though: news and a niche club are not comparable. Current per-category ranges live in the metrics reference.

Common questions

Why do small channels reach a larger share of their audience? In a small channel almost everyone subscribed recently and deliberately. A large one accumulates people who subscribed a year ago, lost interest, and never bothered to leave. A subscriber stays in the counter forever; a reader does not.

What engagement rate is normal on Telegram? It depends on size and topic. As a rough guide from our data: 20–50% for channels under 5,000 subscribers, 8–12% above 100,000. Below 2% means the subscriptions are there and the readers are not.

Can view counts be trusted? Views are easier to inflate than subscribers, so they mean little on their own. The check is whether the numbers agree with each other: views against subscribers, views against posting frequency, reach against mentions elsewhere.

What if the channel is private? A private channel exposes no public feed, and no external measurement of it exists. Any service claiming otherwise is showing you either stale data or invention.

Читайте дальше

A Third of Telegram Links Are Dead: 11,596 Out of 37,683 Checked

We resolved 37,683 Telegram usernames collected from directories, listicles and posts. 30.8% did not open. Here is why t.me links rot, and what to do

Vetting a Telegram Channel Before You Buy Ads: A Ten-Minute Checklist

Seven checks you can run from open data before paying for a placement — reach instead of subscribers, CPM, view spread, cadence, last post date, menti

Telegram Metrics Glossary: ER, Reach, CPM and What Counts as Normal

Subscribers, views, ER, ERR, CPM, posting frequency — each metric with its formula, what it actually measures, and the normal range from our base of 2