How to calculate CPM for seeding in Telegram?

The price of placement is divided by the coverage of one post and multiplied by a thousand. You need to calculate it based on the coverage of the post from the channel statistics, and not on the number of subscribers: the difference between them is often twofold, and calculations based on subscribers systematically underestimate the real cost of the contact.

What to check in a channel before buying advertising?

Post coverage for 30 days, subscriber growth graph, reactions and comments on the last twenty posts, how many advertising publications are in a row. A sharp step in the growth graph means the purchase of bots, views without reactions - that the audience is not alive, three ads in a row - that the feed is burned out.

How much does advertising in a Telegram channel cost?

The placement price is not an indicator. You need to compare channels by CPM: a channel for 30 thousand with a reach of 40 thousand is cheaper than a channel for 10 thousand with a reach of 5 thousand. This is not obvious from the price list, but from the cost of a thousand contacts it is immediately obvious.

How are channel placements different from Telegram Ads?

Telegram Ads is an advertising account with auction and moderation. Seeding is a direct purchase of a post from the channel owner. Sowing has higher audience trust and a freer format, but there is no precise targeting: you buy the entire channel, no matter what it is.

How to calculate the subscriber price from seeding?

The placement price is divided by the number of subscribers who came. To count before placement, take a forecast: reach is multiplied by conversion to transition, then by conversion of transition to subscription. The first odds are an estimate; after the first wave they are replaced by their own.

Who is responsible for advertising labeling and erid?

This is recorded in writing before payment. They remember about marking at the last moment, when the post has already been agreed upon, and this regularly disrupts the output. If neither party is ready to take it upon itself, it is better to refuse the placement.

How many channels should I take in one wave?

Five to eight for the first wave. Anything less and you won’t have enough statistics to distinguish a good channel from a random hit. More - there is not enough attention to remove the fact for each within 24 hours after release.

All articles

SMMSeptember 5, 20269 min

Media plan for seeding in Telegram 2026: CPM, channel verification and subscriber price

You need to count crops based on the coverage of a post, and not on the number of subscribers - the difference is often twofold. I’m going over the calculation of CPM, the checklist for checking the channel before payment, marking placements and how to reduce the wave so that the second one is cheaper.

Sowing is the easiest channel to implement and the most opaque to evaluate. You can coordinate a post in an hour, but you can understand whether it was expensive or cheap only if you count it according to the correct value. The right value here is not price or subscribers.

Advertising is shown not to channel subscribers, but to those who opened the post. Everything that is counted from the number of subscribers is counted from the wrong thing.

1. Post reach vs. number of followers

The channel’s price is almost always justified by subscribers: “fifty thousand, so thirty thousand rubles per post.” But the publication is opened by part of the audience, and this part is very different for different channels - from a fifth to a half.

What does this give in practice:

Channel	Subscribers	Post coverage	Price	CPM
A	50 000	8 000	30 000 ₽	3 750 ₽
B	20 000	11 000	18 000 ₽	1 636 ₽

Channel A looks more solid and costs more. At the same time, a contact in it costs more than twice as much. This is not visible from the price list, but from CPM it is immediately visible. The numbers in the table are arbitrary, but the ratio is typical: a large channel with low coverage almost always loses to an average live one.

2. What to watch before paying

Eight points, each covering one way the channel could end up empty.

What are we watching?	Red flag
Post coverage in 30 days	Shows only the number of subscribers
Subscriber growth chart	A step in one day - purchasing
Reactions and comments	There are views, no reactions
Geo and age of the audience	Refusal to show statistics
History of advertising in the feed	Three advertising posts in a row
Price to coverage	CPM twice the market without explanation
Release time and consolidation	Ready to deliver at three in the morning
Marking and erid	Nobody takes it upon themselves

The most telling point is the second one. Organic channel growth looks like a sloping line, purchased channel growth looks like a step. One glance at the schedule saves your placement budget.

3. Marking: without it the wave is useless

Eight placements without markup result in one line in the analytics: “transfers from Telegram, 640.” What of this the most expensive channel gave, and what the cheapest is unknown, and the second wave is planned blindly.

Minimum: own UTM for each placement. If you are directing to a bot, enter your deeplink with the start parameter. The tag is placed before the post is approved, and not after: you cannot redo what has already been published.

About how to read these transitions later in reports - in the analysis marketing analysts, and about combining sources - in the material about attribution.

4. Forecast before placement

To compare channels not only by CPM, but also by the expected price of a subscriber, two coefficients are needed: the share of those who opened the post who followed the link, and the share of those who clicked and stayed.

For the first wave, these coefficients are taken conservative - just so that the channels can be lined up in one row. They have no predictive power, and should not be treated as a promise. Then they are replaced by our own, calculated from the fact sheet, and from the second wave the forecast begins to work.

B crop media plan template both coefficients are displayed in separate cells at the bottom of the sheet - specifically so that they are visible and easy to replace.

5. The fact is removed on the first day

Telegram does not store a convenient history for a specific post: after two days, the statistics are blurred by the following publications, and some of the transitions can no longer be distinguished from organic ones.

Therefore, for each placement, four numbers are recorded during the day: coverage, transitions, subscriptions, and the actual price of the subscriber. Plus the “what was noticed” column - there are things that are not visible in the numbers: the post came out at the wrong time, the text was wrong, the channel placed it between two other advertisements.

6. Wave mixing

At the end of the wave, five values are calculated: total budget, total reach, conversions, subscriptions and average subscriber price. Then the channels are divided into three groups: for repeat, in question, we don’t take any more.

It is this list that is the main result of the wave, not subscribers. Subscribers will end, but understanding which channels work will remain and will reduce the cost of subsequent purchases.

7. Summary

Count from the reach of the post, not from subscribers. Go through the checklist before payment, not after. Mark each placement with a separate label. Remove the fact on the first day. Recalculate the coefficients based on your data.

Five rules, each of which costs money if you skip it. Ready form with all calculations - media plan for sponsored placements in Telegram.

Send me the list of channels you are considering, I’ll look at the price for coverage and tell you where the overpayment is: Telegram or form.