Keeping remarketing lists exclusive within the same campaign/ad group is no easy task. Without a way to prevent overlap, things get muddled: performance, bidding, audiences, etc.

It’s possible to use higher bid modifiers on the more granular lists, but that may not be the optimal option performance-wise. Making separate campaigns or ad groups for specific audiences is also a potential, but that takes up a ton of time. Luckily, custom combination audiences add the needed flexibility to remarketing lists to combat this problem – and they do a lot more.

You may want to use custom combination audiences if you want do any of the following:

  • Combine multiple audiences in a complex way
  • Make specific groups of audiences mutually exclusive
  • Improve the granularity and targeting of your RLSAs

A Few Notes Before You Begin

Custom combination audiences chain together audiences that you have already created using logical operators (explained later in the article). Let’s say you want to create the following audience:

  • anyone who has visited either the checkout page or the email signup confirmation page in the last seven days;
  • and exclude anyone who visited the email signup confirmation page in the last three days

For this example, you would need to have three audiences already built out:

  1. Visitors of the checkout page for the last seven days
  2. Visitors of the email sign-up confirmation page for the last seven days
  3. Visitors of the email sign-up confirmation page for the last three days

Another thing to keep in mind about custom combination audiences is that Google does not make any Similar Audiences for them. Google only makes Similar Audiences for the root audiences.

How to Make Custom Combinations

After you’ve created the requisite root audiences, go to the audience section in the shared library in the Google UI:

Click on the add Remarketing List à Website Visitors as you would normally.

Afterwards, you will find the option to make custom combinations under “Who to add to your list”:

Finally, you should see a menu that looks something like this:

In the screenshot above, I clicked on the “Add Another” button twice to create three buckets of audiences to explain the various options. Each of the grey rectangles that have a logical operator (OR, AND, none) and a “Select Audiences” button are a bucket of audiences. A user must meet the requirements in each bucket to qualify as a member of your custom audience.

There are three types of logical operators used to chain together audiences within one bucket:

  • Any of these audiences (OR)
    • The user needs to belong to at least one audience within this bucket
    • Falling into more than audiences within this bucket does not change the outcome
  • Each of these audiences (AND)
    • The user needs to be a member in every audience within this bucket
  • None of these audiences
    • This is rather self-explanatory. The user cannot be a part of any the selected audience lists

A Basic Example

For one of my clients, I used custom combinations to ensure that the site visitor audiences that are broken out by days are mutually exclusive. Below is what the custom combination would look like for an audience that targets people who visited the site between 3 and 7 days ago:

I selected Site Visitors – 7 Days to be the audience I want to target, which includes anyone who has visited the site in the past seven days. I then excluded anyone in the Site Visitors – 3 Days. This results in a custom audience that targets anyone who visited the site from 3 to 7 days ago. I repeated this process several times with different time buckets to create a graduating scale all the way up to the 540-day limit.

Final Thoughts

While the example above is very rudimentary, you can get creative with the audiences you create with custom combinations. For many clients, audience size will become a limiting factor before you run out of options with custom combinations, which allows you to squeeze as much value as possible from your RLSAs.

If you have any questions regarding the setup for custom combination audiences, feel free to leave a comment!

Leave a Comment

Brandan Hummell
Brandan joined 3Q Digital in April 2017. He attended the University of Virginia and received a B.A. in English & M.S. in Commerce with a concentration in Marketing and Management; he decided to work for Merkle as a paid search analyst because he loved how the role combined account performance analysis and testing with the management of client relationships. He is inherently detail-oriented and enjoys solving problems, so he eventually joined the SEM technical support team in addition to his analyst responsibilities. In his free time, you will find him cooking, homebrewing, snowboarding, and (in the warmer months of the year) riding his motorcycle.