Telegram Engagement Rate — How to Calculate ER and What Counts as Normal
ER — engagement rate — answers the question that neither subscriber count nor raw reach can: is the channel actually being read? Subscribers can be bought, views can be botted, but nobody has yet figured out how to make thousands of real people consistently react to posts. That's why experienced advertisers look at ER before they look at size.
The catch: "ER" means different things in different tools. Let's take both formulas apart — and see what each one is for.
Formula one: ER by reach (ERR)
ERR = average post reach ÷ subscribers × 100%
This is the share of the audience that actually sees the posts. It's often called simply ER, which adds to the confusion — the precise term is ERR, engagement by reach.
These benchmarks aren't guesswork — we computed them from our own catalogue: 799 channels with measured reach, August 2026.
| Channel size | Median ERR | Half of channels fall within |
|---|---|---|
| under 1,000 subscribers | 24% | 19–41% |
| 1,000 – 10,000 | 16% | 10–29% |
| 10,000 – 100,000 | 11% | 6–20% |
| over 100,000 | 10% | 4–18% |
⚠️ A caveat about the sample: these are channels in our catalogue, predominantly Russian-language. Treat the figures as an honest benchmark for that slice of the market, not as the truth about all of Telegram.
Two takeaways. First: ERR falls as a channel grows, and that's normal. Early subscribers cool off, some people leave Telegram altogether, and the channel's engine — fresh, interested readers — is always a minority. A 100k channel with 12% ERR is a solid mid-tier performer; demanding the 40% of a five-hundred-subscriber channel from it is pointless.
Second, and more uncomfortable: real numbers are lower than the folklore says. Articles about Telegram routinely call 30–40% normal for an average channel — by our measurements that's the top quartile of the 1k–10k group, not its middle. If your 5,000-subscriber channel sits at 12%, you're not failing; you're slightly above the median.
Third: anomalies in either direction are suspicious. An ERR of 3% is a dead base — the channel was once inflated with bots or off-target ads. But a 95% ERR on a 50k channel is a flag too: live audiences don't read like that, botted views do. In our sample, almost nothing above a thousand subscribers exceeds 60%.
Formula two: ER by reactions
Reaction ER = (reactions + comments + forwards) ÷ post reach × 100%
This is the share of viewers who responded to the post. The numbers here are an order of magnitude smaller, and that's fine: of the people who see a post, only a minority ever reacts.
Benchmarks: 1–3% is an ordinary living channel; 5%+ is a very warm audience (small expert channels, tight communities); consistently under 0.5% means people scroll without reading.
⚠️ Those three figures are established market practice, not our measurement: we measure reach, we don't collect reactions. Unlike the table above — that one is computed from our own data, and we stand behind it.
Counting reaction ER against subscribers instead of reach is a common mistake that renders the number meaningless: it blends "how many people see" with "how those who see react" into one figure that answers neither question.
How to calculate your ER by hand
- Take your last 10–15 posts. Drop the outliers: one viral post and one pinned-for-a-month post will wreck the average.
- For each, record the views 48 hours after publication — beyond that a post collects random traffic that has nothing to do with engagement.
- Divide average reach by subscribers — that's your ERR.
- Add up reactions, comments and forwards, divide by reach — that's your reaction ER.
Channel cards on GramDirectory compute ER exactly this way: from public posts measured by our own crawl, with the measurement date shown next to the number. Channel owners cannot influence those values — which makes them a useful independent cross-check against a seller's screenshots.
What to do if your ER is low
Low ER is a symptom, not a diagnosis. The usual causes, in order of frequency:
- Dead weight in the base. The channel grew through giveaways or cheap ads — people came for the prize and stayed as numbers. Only time and content fix this; purging bots is cosmetic.
- Posting too often. Telegram doesn't push every post to everyone — for muted subscribers your channel lives in the archive. More noise, more mutes. Fewer but denser posts almost always win.
- Content gives no reason to react. News summaries don't get comments. A position, a question, a debatable claim — those do. If you want reactions, posts must invite them, not merely inform.
- The wrong audience was recruited. Cross-promo with an off-topic channel brings people indifferent to your content. On choosing partners by audience rather than size — see our cross-promotion guide.
In short
ERR (reach ÷ subscribers) shows whether the base is alive; reaction ER (reactions ÷ reach) shows whether the content lands. Compute both from 48-hour post data with outliers removed. A falling ERR on a growing channel is physics, not failure. And to see what ER is genuinely normal in your topic, compare against your neighbours in the niche catalogue.