Data quality

Google Ads conversion tracking: what it counts, and why it differs from your store

Google Ads conversion tracking is the mechanism by which Google Ads learns that an ad click led to something you value, such as an order or a lead, and counts it against the campaign.

About the Google Ads connectorAbout the Shopify connector

The count is Google's, built on Google's rules, and it will not match the orders in your store. That is expected.

How it's counted

A conversion action is a definition of the event you want counted, such as a purchase or a form submission. Each one has a value (fixed, or taken from the order) and a counting rule (once per click, or every time it happens).

There are two routes for the event to reach Google. The tag route puts a piece of Google's code on your site; when a visitor who earlier clicked an ad reaches the thank-you page or fires the purchase event, the tag reports it. The import route sends conversions in from somewhere else, such as your analytics property or a sales system, so that events Google's tag didn't see, including offline ones, can still be matched back to the click that preceded them.

Either way, the result is an event on your side joined to a click on Google's side, within a time window.

What it tells you, and what it doesn't

It tells you which campaigns and keywords preceded the events you care about, which is what the bidding system needs in order to spend towards more of them.

It cannot say how many orders your store took, or how many of its conversions the ads caused. The first is a question for your order data. The second is a question for an incrementality test.

Where it goes wrong

The numbers differ from your store's for reasons that are mostly definitional.

The attribution window shifts the dates. A conversion is credited to a click that happened up to a set number of days earlier, so an order on the 3rd can be counted against a click on the 1st. The daily columns in Google Ads and in your store then line up on different days, and the monthly totals drift at the edges.

View-through conversions count a view as a touch. For some campaign types, Google can count a conversion when the person saw an ad but never clicked it, and then bought. Your store never saw the view.

Modelled conversions are estimates. Where consent or browser restrictions stop the tag from observing a conversion directly, Google can estimate how many occurred and add them to the total.

Double counting inflates the figure. Two conversion actions that both fire on the same purchase, or a counting rule set to "every" where "one" was meant, raise the count without any new orders.

When the numbers don't match, check in this order: the date the conversion is attributed to against the date of the order; whether view-through and modelled conversions are included in the column you're reading; which conversion actions are set to count in the main conversions column; and whether the tag fires once per purchase.

How to read it in your account

The Conversions column in your campaign reports is the one that bidding uses, and it includes only the conversion actions you have marked as primary. The All conv. column includes the rest: secondary actions, which are kept for observation only, plus view-through conversions and other special sources. Which actions are primary, with their counting rules and windows, is set per conversion action under Goals, then Conversions, then Summary. Source: Google Ads Help, About primary and secondary conversion actions and Understand your conversion tracking data.

Segmenting a table by Conversion action (the Segment menu, then Conversions, then Conversion action) shows which action produced the count. Reporting by the day of the conversion rather than the day of the click is a separate set of columns, not a segment: Conversions (by conv. time) and its All conv., value and cost companions, available from March 2019. Those bring the daily figures closer to your store's. Source: Google Ads Help, Use segments in your tables and Understand your conversion tracking data.

Related metrics

Cost per acquisition: spend divided by these conversions, inheriting every caveat above.

ROAS: conversion value divided by spend, with the same inheritance.

Conversion rate: conversions divided by clicks.

Keeping it current

A scheduled Fresh Sheets report can keep Google Ads clicks, cost, conversions and conversion value by campaign current in its own tab, and a second report can keep Shopify orders in another. Fresh Sheets replaces each tab's contents on every refresh, so keep the reconciliation between the two on a separate tab.