Google Ads remarketing automation for a large online store ⚡



Let’s ask some large e-commerce projects which way they choose to automate their entire Google Ads remarketing. In most cases, they launch dynamic goods remarketing or dynamic remarketing via Google Shopping campaigns. But what about category or filter pages? In this little article we will spill the beans on how to automate remarketing for the entire web site based on the example of the large car parts online store.

Applicable for:

  • All large e-commerce projects;
  • Ticket sales and booking web sites;
  • Sites with the large amount of pages and different levels of enclosure;
  • Projects that utilize apps with multiple pages, such as Walmart or ASOS).

How did it all begin

Once [shema] team received a request from – a big spare parts online stores with the range of 12 mln. goods. They had already had top-notch search campaigns and they addressed us the following issue: can you launch remarketing to all site pages, where the spare part, make and the model of the car is given? For instance, shocks for Porsche Cayenne. They turned out to have more than 150 000 such pages, so the manual launch would have taken years.

How we set it all up

1/ Having analyzed the whole site structure we found out that there are potentially many more pages for remarketing:

  • [Spare part] + [Make] + [Model] — for instance, “Shocks Porsche Cayenne” (hereinafter: ZMM);
  • [Spare part] + [Make] + [Model] + [Year + Body + Engine] — e.g. , “Shocks Porsche Cayenne 2015 SUV 4.8 gas” (hereinafter: ZMMO);
  • [Spare part] + [Make] — e.g., “Shocks Porsche” (hereinafter: ZM);
  • [Shocks] — e.g., “shocks” (hereinafter: spare part);
  • [Spare part] + [Vendor] — e.g., “Shocks Bilstein” (hereinafter: Spare part vendor);
  • [Search page] — (hereinafter: Search);
  • [Item] — e.g., “Shocks Kayaba 443134” (hereinafter: item).


2/ We’ve developed the logic for the whole remarketing chain and cross-minus lists: 

1/ ZM5/ Item; 6/ Abandoned cart; 2/ ZMM
2/ ZMM5/ Item; 6/ Abandoned cart; 7/ ZMMO
3/ ZV5/ Item; 6/ Abandoned cart; 7/ ZMMO
4/ Search5/ Item; 6/ Abandoned cart
5/ Item6/ Abandoned cart
6/ Abandoned cart Audience category “purchased” 
7/ ZMMO5/ Item; 6/ Abandoned cart
8/ Spare part5/ Item; 6/ Abandoned cart; 7/ ZMMO; 3/ ZV


3/ We’ve written technical specifications on remarketing code integration and feed upload for developers for each page type. 

Example of the ZMM page code (has been installed without GTM):

<script type="text/javascript">

var google_tag_params = {

dynx_itemid: 'REPLACE_WITH_VALUE', /// transfer the id of such pages to the code and the equivalent to the feed 

dynx_itemid2: 'REPLACE_WITH_VALUE', /// leave empty

dynx_pagetype: 'ZMM',

dynx_totalvalue: 'REPLACE_WITH_VALUE', /// the price for the cheapest or most popular item. Leave empty if no items found 



<script type="text/javascript">

/* <![CDATA[ */

var google_conversion_id =; /// Insert remarketing ID from AdWords cabinet 

var google_custom_params = window.google_tag_params;

var google_remarketing_only = true;

/* ]]> */


<script type="text/javascript" src="//">



<div style="display:inline;">

<img height="1" width="1" style="border-style:none;" alt="" src="//"/>





Data feed for ZMM pages:

Data feed for ZMM pages

We performed similar iterations with the rest of the pages. 

4/ And the final step – the launch of automated remarketing in Google Ads.

As it seemed to us, all further actions must have been easy, but it only seemed. Unfortunately, the Google Ads account has a lot of constraints (for example, only 5 million lines can be placed in one feed or the feed weight should not exceed 500 mb). We had to split some feeds into several parts and run in different campaigns. We have placed the largest types of pages in a separate account, as we couldn’t exceed the general limit of elements amount per account. 


Thanks to the joint efforts of great specialists from and our team, we were able to successfully automate this part of the advertisement. This approach allowed us to save a lot of time for PPC specialists, and most importantly – to increase the volume of profit for the online store.


Want to do the same? Contact us:

✉️  |   |   [fill in the form]