The member count is the first thing a follower sees and the easiest thing on the page to purchase. Here is how a bought audience behaves differently from a real one, what you can confirm inside Telegram in two minutes, and where the check stops working.

Why is the member count the first number you see?

Open a Telegram signal channel and the line under the name is a member count. Say it reads ninety thousand. It sits above the pinned message, above the results screenshots, above every word anyone actually wrote. It is the first number a follower reads, and it is the only number on that page that can be ordered like stock: by the thousand, from a public storefront, paid for in minutes.

That is the whole problem in one sentence. The count is not evidence of skill, of discipline, or of a single profitable month. It is evidence that some number of accounts, at some point, pressed Join. It cannot tell you whether those accounts are still there, whether they belong to people, or — the only question a follower is really asking — whether anyone following the calls ended up ahead.

Big does not mean fake. Plenty of large channels are large because they earned it. The point is narrower and harder to argue with: of everything printed on the page, the member count is the item with the lowest cost to fake and the highest position on the screen.

What does an audience actually cost?

Selling counted influence is an established business with a legal record. In October 2019 the US Federal Trade Commission settled its first case against a company for selling fake indicators of social media influence — followers, likes, views and subscribers across Twitter, LinkedIn, YouTube and other platforms. The FTC said the firm had filled more than 58,000 orders for fake Twitter followers alone. The regulator's reasoning is worth borrowing: those numbers are deceptive because the people reading them — advertisers, clients, subscribers — read them as a proxy for merit.

The trade did not stop when one seller did, and it has a large presence on Telegram itself. A paper presented at the 34th USENIX Security Symposium in 2025 catalogued channels selling metric inflation and found 68 channels dedicated to social media manipulation services, pointing at 4,051 distinct websites that boost engagement metrics across 51 platforms. Note the shape of that: the supply is industrial, and it is one search away from anyone who wants a bigger number under their channel name.

The retail price moves week to week and is not worth quoting. The unit is what matters. Audience is sold per thousand, in tiers named after how fast the accounts drain away — which tells you the important thing on its own. The seller expects them to leave, and prices that in.

How does a bought audience behave differently?

Bought accounts do the one thing they were paid to do and nothing else. They join. They do not open the app afterwards. That single behavioural fact produces three tells.

Views go flat while members climb. Telegram's own channel documentation is explicit that "each post in a channel has a view counter", and that views from forwarded copies are counted into the total. Members are a stock; views are a flow. Add ten thousand accounts that never open a post and the stock jumps while the flow does not move at all. The two numbers stop tracking each other, and they are printed on the same screen.

The growth curve steps instead of rising. Real audience growth has causes you can point at: a mention in a bigger channel, a post that travelled, a market event that sent people looking. It arrives as a bump with a story behind it. A purchase arrives as a vertical line — thousands in an hour, on a day when nothing happened, from a channel that published nothing unusual.

Then it bleeds. The cohort erodes on its own. Telegram deletes an account that does not come online for a long stretch — the default in the FAQ is 18 months, adjustable by the user — and accounts caught spamming lose the ability to do it on a schedule of their own. So the classic signature is not a step alone; it is a step followed by a long, steady slide that no content decision explains.

The channel owner can see all of this in detail and you cannot. Telegram's statistics API gives administrators of channels above a certain size a growth graph, a followers graph, a mute graph, views broken down by source and new followers broken down by source. A channel that bought its audience knows exactly which line in that panel it bought. The follower sees one number with none of that behind it.

What can you check inside Telegram in two minutes?

Everything below uses only what a public channel shows anyone who opens it.

Read the view counts on several posts, never one. Scroll ten posts spread over a few weeks and note the eye-icon number under each. One post proves nothing: it may have gone out at 3am, or been a repost, or been the one that travelled. A run of ten gives you a range to hold against the member count.

Compare old posts with new ones. Views accumulate — a post from six months ago has had six months to be seen, so it should not be quieter than last week's. If old and recent posts sit at roughly the same view count while the member count towers over both, the members arrived after the readers stopped.

Look for anything a human has to do. Telegram lets a channel attach a discussion group so that every post carries a comment button. Comments, reactions and forwards all require a person to act. Their complete absence on a channel with a five-figure count is the loudest thing available for free — especially since forwarded copies feed views back into the total, so a channel that genuinely gets passed around shows reach beyond its own membership.

Know what you cannot see. The familiar advice to browse the member list for blank profiles applies to groups, not channels: in practice a channel does not show its subscriber list to the people subscribed to it. If the channel has a linked discussion group, that group is where you can see who is actually present. If it does not, the check is simply unavailable — and its unavailability is not evidence of anything.

For the historical shape, you need an outside archive. Third-party trackers keep subscriber history for public channels — TGStat, catalogued in Bellingcat's open-source research toolkit, records subscriber numbers, subscribers acquired over time and average post reach. That is where a vertical step shows up as a vertical step.

Where does this check stop working?

Be honest about the limits, because a check trusted too far is worse than no check at all.

A private VIP channel shows none of it. No public view counts, no growth history, no comment threads an outsider can read. Whatever you verify, you verify on the free storefront, and the storefront is the part built to be looked at. That is why watching a channel's public calls play out before paying beats any amount of profile-reading — the five checks in how to verify a crypto signal channel all run on the feed, not the follower count.

Views are for sale too, and separately. View bots never have to join anything — they open posts. A channel that bought only members shows the flat-views tell; a channel that bought both does not. The ratio between the two numbers is a question, never a verdict.

There is no threshold that settles it. Analytics services publish a share-of-audience-reached metric, and the typical band shifts with topic, language, channel size and posting frequency. A number outside somebody's rule of thumb is a reason to run the next three checks, not a finding on its own.

A small honest channel and a drained bought one can look identical on one post. Both produce a modest view count against a modest membership. Only the history separates them, which is exactly why one screenshot of one post proves nothing in either direction.

And a large count can be a duplicate rather than a fabrication. A study of 35,382 channels and over 132 million messages, published on arXiv in 2021, identified 83 clone channels republishing another channel's content as their own; their average subscriber count was around 28,500. Those members are real people who joined a real channel — just not the one they thought they were joining. The same pattern shows up in our own archive of clone signal channels running one brand under several names, where following three names is following one operation three times, at three subscriptions.

So what is the count worth?

As a measure of results: nothing, in either direction. Membership measures marketing. It responds to advertising budget, to cross-promotion, to a post that caught, and to a purchase order. It does not respond to whether the last thirty calls worked, and it does not fall when they stop working. Audience size and whether a channel is even still publishing turn out to be unrelated questions — see how many of the channels we index still post.

As a measure of exposure: something real. A large audience means more people acting on the same call at the same moment, which matters for fills on thin pairs and for how a channel behaves when it knows a crowd is watching.

The numbers that carry information are the ones a channel cannot produce by spending money: timestamped entries published before the move, losing trades left visible in the feed, a stop written down at the time rather than remembered afterwards. Those are checkable by anyone, and what a badge does and does not certify is set out in what a verified badge can actually mean. Nothing purchasable moves a rating here — not a listing, not a subscription, not an audience.

The member count is the number printed largest and the one that carries the least. Read it as a marketing budget rather than a track record, and spend your two minutes on the view counts underneath it.

Sources