Bot filtering

Ad clicks attract bots: crawlers following your ads, click fraud, headless browsers, uptime monitors. If they're counted, your visitor numbers inflate, your conversion rates look worse than they are, and junk flows into what gets reported back to the ad platforms.

We filter them for you. There is no setting to turn on and no list to maintain.

How it works

Two independent layers classify every event:

In the browser. Deferred until the browser is idle — so it never delays a page or a first event — the pixel runs a local check for automation: headless browsers, testing frameworks, browser automation tools. The verdict rides along on the events that follow. Detection runs entirely in the browser; no data is sent to any third party.

On our servers. Every event's user agent is classified, independently, against the known crawler population — Googlebot, AhrefsBot, social-preview fetchers, and the rest. Every event means every event: this layer doesn't wait for anything, and nothing a bot does client-side can talk its way past it.

An event flagged by either layer is marked as bot traffic.

What happens to bot traffic

Bot events are kept — you can see them, which matters when you're debugging — but they are quarantined from everything that counts:

• They never become visitors or sessions in your reports.
• They never create or join an identity. A bot that hits your site with a click ID can't claim that click.
• They are never forwarded to ad platforms.

And the part that makes attribution trustworthy regardless: revenue enters Attributely server-side, from Stripe-signed webhooks. A bot can fake a pageview; it cannot fake a payment.

What about traffic from your own site editor?

Page builders and CMS preview panes render your site inside an iframe. The pixel detects that and stays silent — preview traffic never fires events at all, so it doesn't even reach the filter.