Adding a public Telegram channel to a monitor is cheap. Keeping a bad one costs you every day it runs: bot-farmed posts, fake engagement, and a channel that looks active while saying nothing verifiable.
These are the five checks I run before a channel earns a slot in the monitor. They take about fifteen minutes with the public preview surface (t.me/s/<channel>), no account and no API key.
1. Views per post vs subscriber count
Subscriber counts are bought; views are harder to fake at scale. Pull the view counts from the last 20 posts and compare them with the subscriber number shown on the channel.
A channel with six-figure subscribers and three-digit views per post is not reaching its audience. Either the subscribers are synthetic, or the channel is dead and the count is a leftover. Neither belongs in a monitor.
2. Timestamp clustering
Real channels post when something happens, or on a human schedule. Bot-fed channels post on a machine schedule: exact minute marks, fixed intervals, bursts at the same second every day.
Read the timestamps of the last 20 posts. If they cluster at :00, :15, :30, :45 with almost no jitter, treat the channel as a content pump until something else proves otherwise.
3. The repost ring
Copy channels repost each other within minutes. Pull the same story from two or three channels you already trust and check whether this channel's version appears earlier, later, or word-for-word.
If it consistently lands after the trusted sources with identical text, it is a relay, not a source. Relays are fine as redundancy — useless as signal.
4. Admin provenance
The channel's creation date and admin account age are public. A channel created last month, run by an admin account with no history and no contact line, is a disposable asset.
This is not disqualifying by itself. It is a discount: you need corroboration from somewhere else before you cite it.
5. One verifiable fact
Pick the most concrete claim in the last 20 posts — a date, a location, a number, a named document — and try to verify it from a second public source.
If nothing in twenty posts is verifiable, the channel is commentary. Commentary can be useful context; it cannot carry a report.
What the checks produce
A channel that passes all five earns a slot. A channel that fails one gets a note explaining which check failed, so future-you does not re-litigate the decision. A channel that fails three is removed, not muted — a muted bad channel still writes rows and still pollutes the volume alarms.
The output is not a vibe. It is a short scorecard: views ratio, timestamp jitter, repost position, admin age, verifiable-fact count. Five numbers you can compare across channels and defend later.
If you want the full version — eleven checks with the exact thresholds I use, plus the scoring sheet and the wording-diff alerting method — it is in OSINT Starter Pack ($19): https://heyuhe.gumroad.com/l/poddr
And if you would rather hand one channel to me: Single-Channel Audit is one channel, one page, 48 hours, $15 — https://heyuhe.gumroad.com/l/knhcuz