Responsive Search Ads — How To Test & Create Impactful Ads

In June 2022, Google Ads officially sunsetted Expanded Text Ads (ETA) in an effort to have more automation in its ads platform. Most agencies (should have) created Responsive Search Ads months in advance of the new update. However, even after the platform update, advertisers are still able to run their old Expanded Text Ads. We can’t edit them, but we can enable/disable them as well as see their performance data.

Marketers had more control over ETAs, which made them easily testable. Google reclaimed control by enforcing Responsive Search Ads, which left a lot of marketers in the dust about ad testing. Ad testing is a different ball game with RSAs.

What are Responsive Search Ads?

Responsive Search Ads (RSAs) are an automated Google ad type. After supplying Google with up to 15 headlines and 4 descriptions, Google will automatically test the best possible combinations to show to a searcher. Google’s machine learning is powerful, and attempts to improve website traffic by increasing click-through rates (CTRs) and conversion rates (CVRs), as well as reducing wasted ad spend.

They also promise, through the newly enforced ad type, to tailor ad copy based on a searcher’s location (or preferences). Is it the ultimate user-friendly ad type, or is Google slowly reducing the role of the advertiser?

How are Responsive Search Ads different to Expanded Text Ads?

While Responsive Search Ads allow advertisers to enter up to 15 Headlines and 4 Descriptions, ETAs only allowed up to 3 headlines and 3 descriptions. The advertiser would dictate the ad content, and a keyword triggered ads to show up as is, i.e: the copy you entered as Headline 1 would show up as Headline 1 when your ad was shown. The order of the headlines in RSAs, on the other hand, are dictated by Google. If you want to ensure a headline will show up in a certain position, this can only be accomplished by pinning it.

Google played a more passive role with ETAs. Since advertisers had to dictate which ad copy showed up in each position, ETAs were extremely easy to test by swapping one of the headlines in an ad group and testing different messages or even unique selling points (USPs). Tests were controlled, and only one factor was changing at a time.

The Importance of Responsive Search Ads & Why You Need To Create Them (If You Haven’t Already)

Although frustrating to marketers who like control (me!), the imposition of RSAs might not be a bad thing. RSAs are relevant and efficient. Google’s algorithm is strong, so it may be a good idea to trust it. Google will now automatically test different variations of your ad in an effort to find the best performing one for the searcher. It’s also extremely time-consuming to test so many different ads. Rotating 15 Headlines and 4 Descriptions is definitely a tedious task for us advertisers, but is a piece of cake for Google.

Frankly, Google wants Responsive Search Ads, and they want them created properly. Advertisers have tried to cheat the system by pinning headlines 1 to 3, with no other ad combinations, effectively creating an ETA within the RSA. This will result in poor Ad Rank in your search campaigns, and will affect your Click-Through Rate (CTR). Google is the grand master, and trying to circumvent their policies may end up affecting your ad delivery.

Creating RSAs: Best Practices

Creating RSAs are pretty straightforward. The maximum number of headlines is 15. Add 15 Headlines and 4 Descriptions of different lengths into the ad rotation so that Google will use the shorter text for mobile searchers. Make sure each description has a Call-To-Action (CTA) and that unique selling points are made clear in every piece of text. Highlighting your offer, or your main USP is also important, and it could be worth pinning that headline in position 2 or 3. Don’t over-pin headlines in your RSAs as this could affect your ad strength, but that metric is not not worth stressing over. Make sure your golden offer is visible every time your ad is shown.

It’s also important to make sure your Ad has relevance to the search term. Google grades ads on a curve, and even though you put enough money to be at the top of the page, if your ad or landing page is not relevant to the searcher enquiry, your Ad Rank will suffer.

Related Article: Is Your Brand Ready For BFCM? How To Crush It With Google Ads in 2022.

How To Test Responsive Search Ads To Find a Winning Ad

Once you’ve created your initial RSA, it’s always a good idea to test different ad variations to improve ad performance. You can create up to 3 RSAs per ad group. When testing different ad variations, make sure your ad rotation is on “Do not optimize: Rotate ads indefinitely.” Choosing this option will allow each variation to have a fair chance in the real world, without Google prematurely making assumptions about your ad and choosing a winner before it gives the variation a legitimate chance. In general, I like to give tests 2 weeks, depending on budget.

  • Play With Pinning

Pinning a unique selling point or an offer is always a good idea. You want that headline to show every time your ad appears to entice people to click. You can pin multiple headlines in the same position (one of Google’s new features), which still gives the algorithm the control it wants. For example, you can test two different discount angles, one showing a percent off, and another showing a dollar amount off, both headlines pinned at position 2. You can then check the asset report to see how each headline performed. Keep in mind, Pinning only one headline can have a negative effect on performance.

RSA Creation, Google Ads, 2022. Pinning Multiple Headlines In Position 1.
  • Test Different Messages

Another option I like to play with is testing different messages that are still related to the ad group. For example, if I was running ads for a skincare company, one RSA would be centred around the natural ingredients used, highlighting the organic formula that’s suitable for sensitive skin, while another RSA could be centred around how clean and sustainable the packaging is, a win for the environment. Testing two different messages could show you which type of messaging appeals more to your audience. This could also be an indicator of how you can position yourself in the market.

How To Determine a Winning Ad

Determining a winning ad will be dependent on the test carried out. In the asset report can see how many impressions a certain headline, or combination of headlines and descriptions, had. This will be useful if you pinned multiple headlines at a certain position within the same RSA.

Asset Report in Google Ads, 2022. Ad-Level.

As for different messaging, you can compare the two RSAs at the ad level and analyze performance based on your campaign’s KPIs.

For a lead generation campaign, looking at cost/lead is the most important metric. As for eCommerce businesses, Cost Per Acquisition, or Cost/Conv, is generally your North Star. A healthy Conversion Rate (CVR) and Click-Through Rate (CTR) are also indicators of a better performing ad variation.

If you’re running an eCommerce account, it’s a good idea to add ROAS as a custom metric. ROAS is simply Conversion Value / Cost. ROAS will give you an indication of the quality of the conversions you get.

If CPA is low and ROAS is high, this points to a higher Average Order Value (AOV), which could indicate a higher paying audience. If the ROAS is low, this could point the opposite.

It’s up to you whether or not you’re going for higher AOV (you are), and which customers you want to attract. Can you assume that your ad copy attracted the type of audience you wanted? If that’s the case, it may be a good idea to double down on that particular message.

Related Article: How To Generate High Quality Leads for Your Business With Google Ads

What Next? Do You Need To Keep Testing?

If it’s not broken, don’t fix it.

The general rule of thumb was to always have at least two ads competing against each other. This isn’t always the case with RSAs. Although testing ads will push you towards more optimization, if you’ve already ran a test and determined a clear winner, it’s OK to let that ad run solo until you plan another (legitimate) ad copy variation or message to test. If you’re hitting all the KPIs solo, there’s no need to rush to find another RSA test.

That’s a Wrap!

The enforcement of Responsive Search Ads has come with a lot of uncertainty, and a new wave of best practices in ad creation and optimization. Google will always strive towards having more control over our ads, but it’s always best to understand the role a marketer plays in ad creation.

You can always reach me on Twitter or directly at if you have any questions or just want to chat about all things paid media.

%d bloggers like this: