Telegram's built-in statistics show growth, views and a handful of charts — enough to feel informed, not enough to make decisions. Advertisers evaluating your channel, and you evaluating traffic you buy, need deeper numbers. Here are the ten metrics that matter, what "good" looks like, and how to get them.
1. Reach per post (not subscriber count)
The single most honest metric of a channel. Average views in the first 24–48 hours divided by subscribers gives your reach rate. Healthy channels sit between 10% and 40%; below 10% usually means a dead or botted audience. Advertisers pay for reach, not for the subscriber number in the header.
2. View velocity
How views accumulate hour by hour after publishing. An organic curve rises fast in the first 2–3 hours and flattens smoothly. A staircase pattern — sharp vertical jumps at regular intervals — is the signature of purchased views. Check it on your own channel before an advertiser does.
3. Engagement rate (ER)
Reactions plus comments plus forwards divided by views. Telegram audiences are passive compared to Instagram — 1–3% ER on views is normal, 5%+ is excellent. Forwards deserve extra weight: they are the only metric that directly produces organic growth.
4. Join and leave dynamics
Growth is net; the story is gross. A channel gaining 100 and losing 90 subscribers a day has a completely different problem than one gaining 15 and losing 5. Watch leaves in the 48 hours after each post — a spike tells you exactly which content drives people away.
5. Subscriber quality score
Not every subscriber is a person. Quality scoring evaluates each new member: account age, username and photo presence, Premium status, activity patterns, join source. Telegrad computes this automatically for every join and shows the quality distribution per source — so "cheap" traffic that is 60% bots stops looking cheap.
6. Retention by cohort
Of the subscribers who joined in a given week, how many are still in the channel after 7, 30 and 90 days? Retention below ~70% at 30 days means either wrong-audience traffic or content that does not match the promise that brought people in. Cohort comparison across sources is the fastest way to find out which.
7. Cost per subscriber (CPS) — by source
Total spend on a source divided by subscribers it brought — but only counting subscribers who pass quality checks and survive the first week. A $0.30 CPS that collapses to $1.50 after filtering bots and churn is the norm in unmeasured campaigns, and the reason measurement pays for itself.
8. Activity heatmap
When your audience is actually online, by hour and weekday. Posting into the peak instead of your own convenience routinely lifts reach by 20–30% with zero extra spend.
9. Source breakdown
Every subscriber should carry a source label: which campaign, which placement, which video. Without tracked invite links, all growth merges into one anonymous number and optimization is guesswork.
10. Click-to-join conversion
For paid traffic: of the people who clicked your link, how many actually joined? Landing conversion of 30–60% is typical for well-matched audiences. A low rate with high clicks means the creative over-promises or the traffic is junk — anti-fraud filters on the click layer (VPN, proxy, datacenter IPs) clean this up before you pay for it.
Putting it together
Connect your channel to Telegrad and you get all ten metrics in one dashboard: real-time joins and leaves, quality scores, retention cohorts, source analytics and click-level anti-fraud. Free to start, five minutes to set up — and every growth decision after that is made with data instead of hope.