This is the subhead for the blog post
Responsive Search Ads (RSAs) are Google’s new ad unit that dynamically serves various combinations of Headlines and Descriptions and, using machine learning, will optimize delivery based on the top-performing Headline and Description combinations.
In this post, we’ll break down how they work, the pros/cons, and early performance takeaways. Let’s jump in.
How They Work
With RSAs, you input up to 15 headlines and 4 description lines to create countless permutations. Google will show the most relevant message to each user based on various signals (query, time, location, device, etc.).
Pros and Cons of RSAs
First, the good stuff:
- RSAs give you more real estate on the SERP!
- You can show up to 3 Headlines and 2 Description lines; the description line character count has also increased from 80 to 90 characters.
- Google runs automated testing and optimization.
- If these work, you should see higher CTR thanks to increased ad size and more tailored messaging.
Now, the major drawback: there’s not yet any reporting into what combinations perform the best; Google only reports on overall unit performance.
We rolled out a test for a SaaS client and hypothesized that by introducing RSAs, we’d see a lift in overall impressions and a higher CTR.
- Test Details: The test rolled out in our Brand campaign against 7 existing ETAs. Based on Google’s recommended best practices, the ad included 5 distinct headlines, 2 containing the keywords (brand name) and 3 without keywords, all of varying lengths. We also included 2 unique description lines with different CTAs.
- Results: Comparing individual ads, the RSA has had the 2 highest-serving overall, and CTR is higher than that of the top-serving expanded text ad (24% vs. 20%). In aggregate, however, there was negligible increase in impressions or CTR for the overall campaign.
These results are only a few days old (we’ve compiled roughly 12K impressions for RSAs), so we continue to monitor brand results and test across top non-brand Beta ad groups to increase sample size and potential lift from the broader query set. In other words, they’re easy to set up, and Google does the work of testing…so we recommend you throw them into the mix and see how they do.