Testing responsive search ads (RSA) can be tedious, but rewarding.
Google sent an expanded text ad (ETA) the dinosaur way that disappointed many advertisers and marketers.
But the opportunities awaiting RSA test takers are huge, especially when paired with general matching intelligence.
However, if you haven’t been keeping a close eye on these changes—or if running Google Ads campaigns is just one of your job responsibilities—it can be confusing to know what to do.
So today I’m going to walk you through three RSA experiments that you can run the next time you create a search campaign.
But first, let’s make sure you have the context you need to test RSA.
What you need to know about testing responsive search ads
Testing with ETAs was fairly simple.
You can run two (or more) ads with consistent titles and descriptions, compare click-through rates (CTRs), and possibly monitor conversions from the landing page.
This approach was possible because ETAs often:
- They appear the same way to every user, every time.
- Served for the same queries as each other.
- Impressions did not play a significant role in the test.
All other considerations being equal, RSA gets 4 more impressions than ETA.
This requires using a metric called conversions per impression, which you can calculate by multiplying the ad click-through rate by the conversion rate.
In other words:
- RSAs are being submitted to more queries due to their higher ad rank.
- An increase in impressions typically leads to a decrease in conversion rate.
- This same increase also increases overall conversions.
- Marketers should monitor conversions per impression (manual bidding) or conversions within target CPA / return on ad spend (ROAS).
Read: How to A/B Test Responsive Search Ads.
Experiment 1: Fixation vs. Mix and match
The first test you should consider running is to see what effects your install has on your campaigns.
By default, RSAs mix and match titles and descriptions to test and find which combinations people respond to best.
Pinning lets you tell Google which title and description positions should appear exactly where you put them.
Create two identical RSAs. You can fill in as many as you want from 15 headings and four descriptions; Make sure they are identical in both ads.
Then, pin some titles and descriptions to one ad, while leaving the other unmodified.
What will you learn
Studies have shown that RSAs with All items installed It can still get great CTR and conversion rates, but those go up when you’re not installing (or installing sparingly).
Running identical ads with pinned items is a great way to see how much of a gap Google will create between the two, all else being equal.
Experiment 2: Segmentation by message
The following is arguably the most important aspect of the RSA exam – and as a copywriter, it’s my personal favourite.
As ad platforms handle more campaign management through automation, messaging is becoming more important than it has been at some point.
You can create up to three RSAs per ad group, so for this experiment you might want to consider increasing your cap.
Each ad group must have a distinct theme around a set of keywords and use broad match. You might address different personalities, pain points, or even presentations of topics.
What will you learn
This experiment should show you which topics are getting a response and what queries are being shown for.
There is no substitute for knowing what your customers are looking for and want to hear.
And when everyone is using the same core automation at Google, that’s one way you can free up the playing field to gain a competitive edge.
Experiment 3: Pseudo-ETAs with RSA control
While an argument can be made that rebuilding an ETA via install defeats the purpose of RSAs, some advertisers still crave (or need) that control.
Thanks for this approach Chris Ridleywhich defined it as a way to handle the limit of three RSAs per ad group.
Create two pseudo ETAs by installing three titles and two descriptions (without another copy).
The third ad is a true RSA where you leave everything uninstalled, instead of using the space to test new messages.
What will you learn
This experiment allows you to measure the performance (in particular CTR) of the bogus ETAs.
This can be useful for advertisers who need to display certain information at all times, such as those who work in regulated industries.
Read: 4 reasons why you click low volume and what to do.
There is opportunity with RSAs if you use them correctly
Unfortunately, Google still won’t show you the ad that was shown for a search term (the report limits this link to the ad group level).
It also doesn’t show you performance based on titles and descriptions.
While there is a certain element of guesswork involved in decoding RSA performance, you can still structure your campaigns according to the data that Google displays.
Here are some things to keep in mind when testing your next RSA campaign:
- Ad strength is one way that Google provides feedback to advertisers about the quality of its ads. While it’s not something you want to completely rule out, don’t feel the need to make Google-suggested changes to your ads to get higher ad strengths.
- RSAs allow you space for up to 15 headings and four descriptions. While it may be tempting to package your ad and see the results, remember that more items mean more guesswork.
- The structure and messaging of your RSAs account for half of your results. Start early, don’t skip the basics (such as positioning and a good website experience), and work with your customers and stakeholders to set reasonable expectations.
RSAs are one of the best examples of Google’s automation changes that require a change in mindset.
Optimize the surroundings of your campaigns – structure, creative, data – and be patient.
There is room for creative and savvy advertisers to make a big profit.
- How to A/B Test Responsive Search Ads: Step by Step with Tips
- How to improve Google’s Responsive Search Ads (RSAs)
- SEO trends 2022
Featured image: Shaiith/Shutterstock