Seeing Double: The Perks of Similar Audiences

As we all know, Google is constantly introducing new features to help advertisers reach users in better ways. One of their recent developments is Similar Audiences for Search. We were able to test Similar Audiences during beta, and now that it is publicly available, we can’t wait to share our findings. These audiences worked so well for us.

What are Similar Audiences for Search

Similar Audiences for Search are mirrors of your current RLSA audience lists that populate with users who’ve exhibited similar behavior to past buyers and visitors. To clarify, Similar Audiences members have not gone to your website, but they are searching on the SERP with queries in your current keyword set, showing their interest in your product or service.

How are Similar Audiences Created

A Similar Audience List will automatically generate whenever you create an RLSA audience. While the RLSA audience is comprised of users who have visited your site, the Similar Audience list will be made up of users who have not been to your site but exhibit similar behavior to your RLSA audience.

Performance Expectations

The goal of Similar Audiences is to allow for more granular segmentation between broad users and users that are interested in your brand offerings based on their online behavior. By activating Similar Audiences, you can divide your funds and have a more robust targeting strategy in place. Overall, Similar Audiences are expected to perform better than your current non-audience campaigns and slightly below or at par with your RLSA campaigns.

Does Recency Effect Your Similar Audiences?

Depending on your current RLSA audiences, you may have different recency windows applied to the same audience, for example, All Visitors (1 Day) and All Visitors (10 Days). If this is the case, you might be wondering if there is any difference in what version of an RLSA audience you use when applying Similar Audiences. According to Google, regardless of the look-back window associated with your RLSA audience, Google uses a 4-hour span (2 hours before and 2 hours after the event/activity of a general user) to qualify them for an audience. So rather than copying over Similar Audiences for 1 and 10 days, you can reduce your work and only copy over 1 audience to your non-RLSA campaigns.

So, the next question is, “how do the Similar Audiences perform?”

Our Results:

Initially, as with any new program, it took some time for the results to populate. We didn’t see any needle-moving performance straight out of the gate, but now that they have been running for awhile we are very pleased with the results.

The CTR on Similar Audiences increased by 256% compared to non-audience campaign performance

Overall 20% of all non-brand transactions come from the Similar Audiences

Conversion rate increased 7%

When looking at the CPA (cost per acquisition), Similar Audiences had a 40% lower CPA

All in all, we believe the numbers speak for themselves and fully recommend utilizing Similar Audiences to other advertisers.