Are You Taking Advantage of the New Ad Space in Google Ads Search Network?

In the quest for more volume in Google Ads, it’s easy to first look toward GDN (Google Display Network), search partners, and broader keywords. Especially now that Google Search is increasingly less likely to show ads for long-tail keywords with lower search volume.

However, there is still more traffic to capture and an opportunity to regain impressions for your carefully crafted long-tail keywords. According to Google, there is more ad space available for those using the new ad format Responsive Search Ads (RSA). This isn’t about a new placement or network, but rather about search terms.

With RSA, you give Google the ability to use machine learning to assemble the ad they consider most relevant for each search term. With this, Google is confident enough to allow more long-tail traffic. This is volume that handwritten ads don’t quite reach.

What Are Responsive Search Ads (RSA)?

With RSA, you provide 3–15 headlines and 2–4 descriptions, and Google decides the placement and combination using machine learning.

The format in the search results looks the same as regular ads, but Google controls the order of the ad’s components.

RSA is currently in BETA and therefore not shown for all potential searches. The closer we get to a full launch, the higher the share of RSA impressions will be. That’s why it’s important that each ad group also contains regular ads.

Lower Quality?

It’s worth mentioning that the quality of this extended traffic is often lower. This is because some search queries may be only loosely relevant to the ad when Google has more flexibility. This also applies to exact match keywords which, for some time now, haven’t been truly exact but more lenient toward variations.

Many advertisers naturally A/B test RSA against their handwritten ads and quickly dismiss RSA. In such a test, your carefully crafted ad copy should, of course, perform better than Google Ads’ automated ads, making it easy to classify RSA as a low-performing format. But here lies a valuable volume: it may perform somewhat worse, but it’s still traffic you didn’t have before. And more conversions at a lower conversion rate is still better than none at all. For this reason, RSA is worth a closer look.

Why Don’t Regular Ads Show for These Search Terms?

Traditional ads aren’t shown as often for these queries because of lower Ad Rank.
If you’re a skilled advertiser, you’ve written three ads per ad group, giving Google a set number of components to work with. This can easily result in a lower quality score for many searches where another combination would have been better. Remember that quality score is a dynamic value that changes with every search, so the score you see in the interface is too rough to judge an ad’s quality.

With RSA and machine learning, Google instead has the opportunity to create the best ad for each query in real time.

Best Practice

Even though RSA works well and performs better than expected, you shouldn’t abandon handwritten ads. In fact, you must still use them for your most important and best-performing keywords, even after RSA leaves BETA and is shown for all searches. For skilled advertisers, handwritten ads will always perform better in terms of conversion rate. That makes sense since you can tailor them more precisely with a personal tone that directly appeals to your audience. But there’s value in letting RSA handle long-tail advertising and seeing what Google’s machine learning can do with your messaging.

Be thorough in writing strong headlines and descriptions with varied messaging and call-outs. Otherwise, you won’t take advantage of the format and Google’s ability to assemble the best-performing ad.

Evaluate the Data

With broader advertising for more search terms, you also gain more data to analyze for your handwritten ads. However, in Google’s interface, RSA data is somewhat limited. Only impressions for different combinations can be reported and segmented by day of week, time of day, device, and similar. But this doesn’t provide insights into clicks, conversions, and, most importantly, search terms—the metrics we really need.
It is possible to retrieve that data via the Google Ads API and then merge it in Excel to compare RSA and regular ads at the search term level. But that’s an exercise for another post!

As a first step, I recommend looking at the search terms report before and after RSA implementation. From that comparison, you can roughly estimate the character and volume of search queries. Keep in mind, though, that other factors such as seasonality, quality score, and bid changes may affect that data.

In Summary

Whether you’re a sharp SEM specialist or an enthusiastic amateur, you can benefit from and save time with RSA and Google’s powerful machine learning.

As long as you provide strong components in the form of varied headlines, USPs, and descriptions, Google will generate solid results with this format.

For those with limited time, it’s a convenient way to gain traffic volume of acceptable quality.

If you’re a Google Ads specialist, you’ll naturally want more control and detailed data. That data is available but requires some digging. Once you have it, you can set up processes to leverage the format in the best way possible.