← All articles

Affiliate link tracking: why the click counts never match (and what you can prove)

Published By Just Shrink team8 min read

You sold a placement: a sponsored post in your Telegram channel, a link in a video description, a slot in a newsletter. A week later the advertiser says the post brought "about 400 clicks", while you expected far more. Or the reverse: your counter shows 900 and the advertiser asks why only 300 people reached their site.

Neither of you is necessarily lying. Clicks get counted at different points, by different systems, under different rules. This guide explains the usual causes, what a neutral tracking link can settle, and what it can't.

Who this is for

  • Channel owners and creators who sell posts and get asked "how did it perform?"
  • Affiliates and publishers paid per click, lead or sale, who can't see the advertiser's numbers.
  • Media buyers buying placements in Telegram channels who want a figure both sides accept.

The seven usual problems

1. Two counters, two definitions of a click

Your counter records a hit when the short link is opened. The advertiser's tracker (or their web analytics) records something else: a session started, a page loaded, a tag fired. Google says the same about its own products: ad clicks and analytics sessions differ, because a person can click twice in one session, close the page before the tag loads, or come back later without clicking. Between two different companies' systems, a gap is normal.

What to agree on up front: which counter decides payment, and what gap is acceptable.

2. Bots and link previews

When you post a link in Telegram, the app fetches the page to build a preview. So do other chat apps and social sites. Telegram's previewer identifies itself as "TelegramBot (like TwitterBot)". Security scanners, uptime checkers and crawlers add more hits. A plain shortener counts all of these as clicks, which inflates your number before any person has seen the post.

Advertisers usually filter these out on their side. So the honest number to report is human clicks, with bot hits shown separately, not one mixed total.

3. Repeat clicks and unique visitors

One person can open the same link five times: from the post, from a forward, again later on another device. "Clicks" and "unique visitors" are different numbers, and the two sides may be quoting different ones. Say which one you report.

4. UTM tags lost on the way

Advertisers ask you to use a link with UTM parameters so their analytics can attribute the traffic. Things go wrong when:

  • someone in the chain "cleans" the link or copies it without the query string;
  • a redirect chain drops parameters;
  • the link opens in an in-app browser (Telegram, Instagram, others) with its own cookies, so a visitor who later comes back in their main browser looks like a new, "direct" visitor.

The result on the advertiser's side is "we saw almost nothing from your channel", even though people clicked.

5. Redirect chains

Affiliate links often pass through several hops: your short link, the network's tracking domain, the advertiser's tracker, the landing page. Every hop adds load time and a chance for a slow mobile connection to give up. Each system in the chain counts at a different moment, so the numbers get smaller at every hop.

6. Postbacks you can't see

In affiliate marketing, conversions (sign-ups, deposits, sales) are usually reported by the advertiser's system to the affiliate network through a server-to-server postback. The affiliate sees whatever the network shows. If a conversion isn't attributed, or is "scrubbed" as invalid, you have no independent view of it. Disputes about shaving (conversions under-reported to cut payouts) are a well-known complaint in the industry, and a click counter alone can't prove or disprove them.

7. No shared source of truth

Most disputes end as one screenshot against another. Without a figure both sides can open at any time, the side with the bigger budget usually wins.

What a neutral tracking link can prove

A short link that you control, put in the post itself, gives you an independent record of what happened before the advertiser's systems take over.

With Just Shrink, each link records:

  • Visits (humans) and Bot views as separate counters. A hit whose user agent contains "bot", or has no user agent, counts as a bot view; everything else counts as a visit. Detection is approximate: a previewer that doesn't call itself a bot is counted as a visit.
  • Unique visitors, one per browser (cookie on the short-link domain).
  • Countries (from the IP) and referrers (Telegram, Google, X or direct).
  • 7- and 30-day trends and a daily chart, so you can see when a post stopped working.

Stats stay with the link while people keep using it. After 6 months without a human visit, the link and its stats are deleted; bot hits don't count.

And you can share it:

  • A report link for the sponsor. Publish a read-only snapshot of one link or your whole account. The sponsor opens it without an account. You choose when it expires (1, 7 or 30 days, a date, or never), can unpublish it at any time, and can save it as a PDF.

This settles the first three problems: what was clicked, by people or by bots, and how many distinct visitors. It also helps with the fourth. Put the advertiser's full UTM-tagged URL as the destination of your short link, so the parameters live behind the short link and nobody can strip them while copying.

One link’s statistics: visits, unique visitors and bot views, counted separatelyOne link’s statistics: visits, unique visitors and bot views, counted separately
One link in the cabinet: Visits and Bot views counted separately (sample data).
Public shared report for one link: snapshot banner, statistics and visits by dayPublic shared report for one link: snapshot banner, statistics and visits by day
The read-only report a sponsor opens without an account (sample data).
Sponsored post in a Telegram channel
Just Shrink sees
Link preview fetch by TelegramA bot hit, not a person: shown as Bot views
Reader taps the link
Short link redirect (s.gbitcode.com)Counts human visits, unique visitors, bot views, country and referrer
Advertiser sees
Advertiser's tracker / landing page with UTM
The conversionAction: sign-up, install, purchase

Conversion happens here — Just Shrink doesn't see it

Postback / conversion reportInside the advertiser's or affiliate network's systems

Some readers drop off between ③ and ⑤ (in-app browser, slow page) — a common source of click mismatches.

How far a link tracker sees: Just Shrink counts what happens up to the redirect; the conversion happens later, in the advertiser's systems.

A seven-step journey from a sponsored post in a Telegram channel to the conversion report. Steps 2 to 4 are in the group "Just Shrink sees", steps 5 to 7 in the group "Advertiser sees".

What it does NOT solve

Be clear about the limits, with your sponsor and with yourself:

  • It doesn't track conversions. Just Shrink counts clicks on the short link. It doesn't see sign-ups, deposits or sales, and it doesn't receive postbacks. Use the advertiser's or network's conversion tracking for that, and ask for access to the postback or conversion report.
  • It can't see past the redirect. Whether the landing page loaded, or how long the visitor stayed, is on the advertiser's side.
  • Bot detection is approximate. It catches crawlers that identify as bots. It isn't fraud detection, and it won't catch a bot pretending to be a browser.
  • It isn't an ad-labelling tool. It doesn't add or manage ad-disclosure IDs (such as ERID in Russia). If the advertiser needs a marker in the URL, put it in the destination link.
  • It can't count channel subscriptions. Telegram invite links aren't redirects, so a "subscribe to my channel" placement can't be measured with a short link.

A workable setup for sponsored posts

  1. Get the advertiser's final URL, with their UTM tags or affiliate parameters.
  2. Create one short link per placement (per post, per channel, per date). Name it after the deal.
  3. Put the short link in the post, not the long URL.
  4. After the post runs, read Visits and Bot views separately. Report Visits as "human clicks".
  5. Publish a report for that link and send it to the sponsor, with an expiry that covers the payment period.
  6. Ask the sponsor for their sessions or conversions for the same period. Compare the two numbers, not two screenshots.

Checklist before your next placement

  • We agreed which counter is used for payment (yours, theirs, or the network's).
  • We agreed on an acceptable gap between counters, in writing.
  • We agreed on "clicks" vs "unique visitors" vs "sessions".
  • I asked whether their tracker filters bots, and how.
  • The destination URL includes their UTM tags or affiliate ID, and I tested it.
  • There is one tracking link per post or slot, named so I can find it later.
  • I counted the redirect hops and removed the ones nobody needs.
  • For CPA or revenue-share deals, I have access to the conversion or postback report.
  • The report link for the sponsor stays live at least until the invoice is paid.
  • I saved my own copy (PDF) of the report. Idle links are deleted with their stats after 6 months.

Try it on your next post

Send the advertiser's link to @just_shrink_bot in Telegram, or use the web app at jshrink.gbitcode.com/app. You get a short link with human and bot hits counted separately, and a report you can forward. It's free.

Sources (4)

Checked 2026-11-16.