See what each ad set’s traffic does after the click.

What Ads Manager doesn’t show you

Ads Manager counts the click. What happens next is on your page, and most of it never reaches Meta: the visitor who gave up before the page loaded, the data-centre traffic an audience network placement sent, the phone with an ad blocker where your pixel never fired.

Each of those is a paid click that looks like a visit and behaves like nothing. Averaged across a campaign they disappear. Split by ad set they point straight at the one to fix.

How it works

1. Upload your pageA zip, the files, or import it from where it lives now. We check it, approve it, and it goes live on your domain with one DNS record.
2. Tag your ads onceThe Hub gives you one line of URL parameters. Paste it into each ad’s tracking field and Meta fills in the campaign, ad set and ad names by itself.
3. Read it by ad setVisits, CTA taps, leads or Telegram joins, suspicious share and countries, per ad set, for the last 24 hours, 7 days or 30 days.
4. Get told when something slipsThe page stops answering, the pixel goes quiet, or one ad set starts sending suspicious traffic: you hear it on Telegram or push first.

Suspicious, and why we say it that way

No tool can prove a visit is fake, so the Hub doesn’t pretend to. It flags a visit as suspicious when it shows one of three plain signals, and shows you which one.

Two kinds of traffic are deliberately left alone. iPhone users on iCloud Private Relay arrive through large network providers and are real people, so they’re never counted as data-centre traffic. Meta’s own review and preview visits are counted apart and left out of every number, so they don’t dilute your rates.

Blocked pixels, recovered

On a share of phones the Meta pixel never loads: ad blockers, strict browsers, privacy settings. Those visits still cost you, and Meta never hears about them. The Hub measures that share on every page and, with your Conversions API token saved, sends what was lost from our server.

What you see

What we keep, and for how long

No cookies of ours. No IP addresses stored: only a salted fingerprint of network and browser, used to spot the same device coming back. Visit records are kept for 45 days, then deleted. The only thing sent anywhere is what goes to your own pixel.

The limits, plainly

It reads pages we host, because the tracking is built into them. Ads without the URL parameters still count, filed under “No ad set tag”. Suspicious is a signal, not a verdict: a real person on a work VPN can land in it, and a careful bot on a phone network can miss it. Some phones don’t report every timing, so speed is measured on the visits that do. Recovering blocked events needs your Conversions API token.

Questions about Page Fellow