Churn in Paid Telegram Channels: How to Measure It, and What Good Looks Like

Churn in Paid Telegram Channels: How to Measure It, and What Good Looks Like

Most people running a paid Telegram channel have no idea what their churn rate is. They know roughly how many members they have and whether the number went up.

That is not enough to run a subscription business, and Telegram adds a specific complication that makes the standard advice misleading.

First, the arithmetic

Monthly churn = members lost during the month ÷ members at the start of the month.

If you started March with 400 subscribers and 28 left, churn was 7%.

Two rules that people get wrong constantly:

Do not net off new signups. Losing 28 and gaining 40 is not "12% growth and no churn." It is 7% churn and 10% growth, and they are separate problems with separate fixes. Netting them hides a leak until growth stalls and the leak is all that's left.

Compound it before you feel relaxed. 7% monthly churn sounds small. It means you lose 58% of a cohort in a year. At 5% you lose 46%. Monthly churn is the most flattering way to express the number, which is why everyone quotes it that way.

Published benchmarks

These are industry-wide subscription figures, not Telegram-specific. No credible Telegram-specific dataset is published, and anyone quoting one should be asked where it came from.

Rough interpretation for a paid channel: under 5% is good, 5–8% is normal, above 10% means something is broken.

Treat these as orientation, not a target. Your niche, price point and audience matter more than any benchmark.

The Telegram-specific problem

Here is where the standard subscription playbook breaks down.

Split churn into two kinds:

• Voluntary — someone decided to leave
• Involuntary — the payment failed

In conventional subscription businesses, involuntary churn runs at 20–40% of total churn, and roughly 10–15% of recurring payments fail on first attempt — expired cards, insufficient funds, bank declines.

Telegram Star subscriptions renew by automatically debiting the subscriber's Star balance every 30 days. There is no card, no expiry date, no bank.

There is something arguably worse: a prepaid balance that is usually near zero.

A card generally has money behind it. A Star balance is a wallet most users top up for a specific purchase and then run down. When renewal day arrives, the question is not whether their bank approves — it is whether they happened to have Stars sitting idle.

We would expect involuntary churn on Star subscriptions to run structurally higher than the card-based benchmarks above, because the funding mechanism is prepaid rather than credit. That is a reasoned inference from how the mechanism works, not a measured figure — but if your churn looks worse than the 5–8% benchmark, this is the first thing to check, and it is the cheapest thing to fix.

It also means a large share of your churn may be from people who still want your channel. They did not decide to leave. They simply had 40 Stars when they needed 500. That is a solvable problem in a way that genuine cancellation is not — see handling failed renewals.

What to actually track

Four numbers, monthly:

1. Members at period start
2. Members lost — split into failed-renewal and deliberate-cancellation if your platform distinguishes them
3. Members gained
4. Involuntary share — failed renewals ÷ total lost

That fourth number is the one that tells you what to do. High involuntary share means a payments problem: warn earlier, encourage bigger top-ups, add grace periods. High voluntary share means a product problem, and no amount of dunning will fix it.

If you can, add cohort retention — of everyone who joined in January, what percentage remain after 1, 3 and 6 months. Aggregate churn is dragged around by growth; cohorts tell you whether the product is actually holding people.

Where the Telegram data lives

At the API level, channel admins can see the end date of a subscriber's current period (subscription_until_date on the participant record). That gives you the renewal boundary for every member, which is what you need to distinguish "left at renewal" from "left mid-period."

Note a structural constraint that shapes everything here: Telegram's native Star subscriptions only support a 30-day period. The Bot API's createChatSubscriptionInviteLink requires subscription_period to be exactly 2592000 seconds. There is no native annual plan, which means:

• Every subscriber faces a renewal decision twelve times a year
• You cannot use annual billing to reduce churn, the standard lever in every other subscription business
• Your churn will look worse than an equivalent business selling annual plans, because it is measured against twelve renewal events rather than one

That is not your product failing. It is the billing cadence, and it is imposed.

The benchmark that matters

The number that would actually help here — aggregate churn across real paid Telegram channels — is not published by anyone, and that includes us. Industry subscription benchmarks are everywhere; Telegram-specific ones are nowhere.

So compare against yourself. Track the four numbers above monthly, watch the involuntary share, and judge this month against your last three rather than against a figure you have no way to verify. That is a less satisfying answer than a benchmark, and a more honest one.

What to do with the number

Above 10% monthly: find out whether it is voluntary or involuntary before changing anything. The fixes are completely different and doing the wrong one wastes a month.

5–8%: normal. Improve onboarding and the first 30 days — first-period churn is usually the largest single cohort.

Under 5%: you have retention. Spend on acquisition.

No idea: start counting this month. Four numbers, once a month, is enough to run the business properly.

Benchmark figures are industry-wide subscription data from published 2026 sources, not Telegram-specific. API constraints verified against Telegram's Bot API documentation, September 2026.