Google Shopping (Google Product Search) can be a very powerful source of traffic and conversions for online retailers. However, analysing the performance can be very tricky as currently there does not seem to be a solid integration between Google Shopping and Google Analytics (GA). As such, to get the required data accuracy for analysis and decision-making purposes, customised GA tracking solutions need to be used.

Currently, there are two GA tracking customisation methods that seem to be advocated by analytics users to track Google Shopping traffic:

  1. URL filtration: using advanced filters imposed on either the referral URL or the request URI to rewrite the Source or Medium information to differentiate this traffic from Google / Organic traffic.
  2. UTM variables: using UTM variables appended to the back of the Google shopping feeds which power the Google Shopping listing.

We will assess each below and demonstrate why the latter is better than the former.

Method 1: URL Filtration

This method is the simplest as it only requires adding filters onto the GA profiles where the traffic is to be tracked and analysed.

However, due to the different scenarios below, this method does not seem to provide enough reliability to guarantee data integrity.

Currently, there are three different ways for a user to reach Google Shopping search results:

Tracking Google Shopping with Google Analytics
Method 1: Go to, click on Shopping link at the top of the page, and search for “red long dress” Method 2: Go to, search for “red long dress”, and click on Shopping link at the top of the page Method 3: Go to, search for “red long dress”, and click on Shopping link on the left

It turns out that the URLs of the Search Engine Results Page (SERP) and the URIs to the actual product (on the retailer’s websites) vary depending on how the user reaches the SERP:

Not only that, but the lack of distinct characteristics on these three SERP URLs (as the “referring page”) is making it hard to differentiate this from a normal organic search SERP.

The URL filtration method would rely on detecting certain pattern in the referral URL (i.e., the SERP URL; e.g., matching “”) or in the request URI (i.e., the result URL which users click on; e.g., matching “”)

As shown in the above tables, should we want to track using URL filters, the referral URL and the request URI are inconsistent from one method to the other. And since there is no clear way to establish what the pattern is, tracking using URL filters may well lead to a very inaccurate or incomplete data.

Due to this weakness, the URL filtering tracking method is thus not recommended.

Method 2: UTM Variables

Another method to track Google Shopping traffic is by implementing UTM variables on the URLs in the product feed.

This is as simple as appending the following line to the back of every product page URL in the feed:


  • the utm_campaign value can be changed to any arbitrary identifier value
  • the use of utm_term and utm_content is discouraged as this would threaten the integrity of the search keyword data

Thus, the #1 result example URLs tagged with UTM variables are provided in the table below. Note: The appearance changes are due to percent-encoding done by Google SERP processing.

When implemented properly, the landing page URL would correctly show the three UTM variables upon the user landing. Furthermore, using tools such as Firebug to audit the __utm.gif would show correct Source/Medium information (Google/Shop) while retaining the keyword data intact.

This method is foolproof for any referral URL and request URI variations, as it only relies on the final landing page URL which is specified in the XML product feed and is 100% controllable by the vendor. Thus using UTM variables, and not profile filters, is the recommended method.
Author’s Note: Google Shopping is currently unavailable in New Zealand.