There's a specific feeling you get when you're looking at a channel before buying ad placement — decent subscriber count, post views in the right ballpark, but something doesn't add up. Reactions are almost zero. The growth graph has those vertical spikes with no explanation. You can't quite articulate it, but the numbers feel constructed. That instinct is usually correct.
Reading the numbers without any tools
The most reliable single metric is ERR — average views of the last 10 posts divided by subscriber count, times 100. A 50,000-subscriber channel should sit roughly at 10–20%. Significantly below that, start looking harder.
Reactions are your next check. Real readers click buttons — not always, but consistently. A healthy channel converts somewhere between 0.5% and 3% of views into reactions. A post with 80,000 views and 40 reactions isn't a real audience; it's a number on a screen. Forwards work the same way: content that genuinely resonates gets shared to chats, saved to personal history. Zero forwards alongside thousands of views is telling.
Pull up the subscriber chart and look at its shape. Organic growth has a certain randomness — peaks after a viral post, slow weeks, gradual drifts up or down. Bot purchases produce clean vertical spikes: the subscriber count jumps by 3,000 in a single day with no visible cause, then plateaus or drops. Sometimes you see the sawtooth — spike, then a decline wave a few weeks later when the service purges old accounts.
If you can access the member list, browse a random sample from somewhere in the middle — not the first 50, those tend to be older genuine subscribers. Count the proportion with no avatar, no username, no bio, last seen "a long time ago." Five percent is normal. Twenty percent is worth investigating.
What manual checks can't catch
Manual inspection catches the obvious cases. It won't catch aged bot accounts — accounts bought years ago on marketplaces, with some activity history and maybe a profile photo. These pass casual inspection cleanly. It also won't surface temporary bot rentals, where accounts join, boost numbers, and leave before anyone notices. You can get to high confidence by triangulating several signals together, but high confidence isn't a verified number.
How a full audit works
TGuard's audit cross-references your channel's member list against a database of over 10 million known bot accounts — built up over three years of monitoring 12,000+ channels. This isn't manual comparison; it runs automatically against the full list.
Beyond known bots, the audit separately identifies dead accounts (last seen a long time ago, no recent activity) and deleted accounts (people who had the account when they joined but have since deleted it). These are distinct things with different implications — a deleted account is simply gone, while a dead account may still technically inflate your view count through Telegram's preview systems.
What the report gives you
The output is a cleanness score from 0 to 100, with a breakdown showing exact counts of bots, dead accounts, deleted accounts, and clean subscribers alongside the number of members actually scanned. Most channels have never run this check, which means showing a 90+ cleanness score is a genuine differentiator when pitching advertisers.
When to actually run one
Before buying advertising is the obvious answer. But before selling it matters just as much — a transparent audience quality report shifts the negotiation in your favor. Channel acquisitions are another strong case: a channel's value is mostly its audience, not its subscriber count, and without an audit you're buying a number rather than a business.
One situation people consistently overlook: unexplained growth. If your channel gained several thousand subscribers in a week with no obvious cause — no viral post, no mention in a bigger channel, no campaign — someone may have inflated you deliberately. It happens as a competitor sabotage move, timed to damage your analytics before a deal goes through.
Frequently Asked Questions
Key signs: ERR significantly below the normal range for the channel's size, sudden subscriber spikes without visible external promotion, large numbers of blank profiles in the member list, and mismatched view and reaction counts.
Manual checks include browsing the member list for blank profiles, reviewing the subscriber growth chart, and calculating ERR manually. However manual checks cannot detect sophisticated bot networks with aged accounts.
TGuard's audit scans channel members against a database of over 10 million known bots and identifies dead and deleted accounts. The result is a cleanness score from 0 to 100 with a breakdown of bot, dead, and deleted counts.
The audit is just the starting point: once the check is complete, TGuard doesn't stop there — it continuously blocks new bots at the moment of every join attempt. The clean state of your channel is maintained automatically going forward, not just documented in a one-time report.