A Tracking Watchdog for Stores Whose Ads Run on AI

A Tracking Watchdog for Stores Whose Ads Run on AI

In ad requests, Google Analytics is named more often than Facebook Ads or Google Ads, and one request in four is about tracking. The AI now buys the ads. The product opening is checking what it learns from.

Meta's Advantage+ and Google's Performance Max now choose the audience, the placement and the bid. Both learn from one thing: the purchases a store reports back through the Conversions API and conversion tracking.

The AI is only as good as the purchases it can see.

The most named product in ad requests is not an ad platform

When businesses ask for help with paid ads and name a product, Google Analytics comes up more often than Facebook Ads, and far more often than Google Ads. Tag Manager is on the list too.

Products named in requests about paid ads

Requests about paid advertising that name at least one product.

About one ad request in four is about the measurement itself: the pixel, the Conversions API, GA4, the tags. That share is climbing.

A broken pixel does not stop the spend. It teaches the AI the wrong lesson.

Tracking breaks quietly

A theme update drops a tag. A new checkout skips the purchase event. A cookie banner blocks the pixel for half the visitors. The ads keep running and the dashboards still show numbers.

The campaign now optimises toward whatever events still arrive. The store finds out weeks later, when the cost per sale has drifted and nobody can say why.

The tools that exist set it up once

Server-side tracking is not new. Elevar, Littledata and Stape install it, mostly for Shopify, and the best known start at around $200 a month. They solve the setup.

The businesses asking look different:

Who asks for tracking help
1 in 4ad requests are about trackingup from 1 in 5 earlier
1 in 7tracking requests come from a Shopify storemost mention no store platform at all

Requests about paid advertising.

Most run WooCommerce, Wix, Webflow or a checkout of their own, and spend far less than an Elevar plan assumes.

What to build

  • A daily tracking watchdog. It compares the orders in the store with the purchases Meta, Google and GA4 received, and tells the owner the day they drift apart.
  • Server-side tracking for every other store. The Conversions API and enhanced conversions for WooCommerce, Wix, Webflow and custom checkouts, priced for a store spending a few thousand a month.
  • A signal report in plain words. One page that explains what the ad platform can and cannot see, and the one fix that matters most this week.

The platforms automated the buying.

Build the part that checks what they learn from.