This is a guest post from Charlotte Haab, Account Lead at
A few months back I gave you the scoop on Google’s latest ad type: responsive search ads. How have these ads been performing since their release?
Refresher: The Lowdown on RSAs
Responsive ads essentially mix and match from a variety of pre-written ad assets. In other words, you can write up to 15 headline variations and four description variations, and Google will determine which combinations are most relevant to each individual auction. It’s also worth noting that each ad can show up to three headlines in the SERP, with extended 90-character descriptions.
Early Positive Results
At this early stage, the consensus is mixed. Google’s research shows that advertisers can get up to 15% more clicks or conversions by adding a responsive search ad to their ad groups.
Our own research shows that some clients saw nice improvements to click-through rate (CTR) and conversion rate (CVR) using RSAs, while others didn’t.
Why was this the case?
For one eCommerce client, the challenge was expanding branded keyword efforts when we were already at 100% impression share. With responsive search ads, we were able to grab about 10% more impressions that we weren’t getting previously.
One thing to keep in mind, in this instance, is that RSAs had just been released. A few weeks after their release, Google announced updates to standard ads that included three headlines, and two 90-character descriptions. This inherently puts less tailored RSAs at a disadvantage when comparing.
For this same retailer it’s also worth noting we saw a nearly 15% increase in CTR and a 12% increase in CVR. That’s because RSAs give Google the freedom to use their algorithms to decide which combination of copy is most likely to drive a conversion for each individual auction. The more conversion volume your account has, the better Google can learn what works for who. As a trade off, there’s a limitation to how query-specific your ads can be.
If your account has ample conversion volume and uses largely generic or branded copy—you can probably expect some improvements to performance and impression share from implementing RSAs.
In a contrary example, another similarly sized eCommerce retail client implemented RSAs and saw the opposite happen. CTR dropped by about 15%. While this account had similar levels of traffic, they sold a more specialized product and had a longer time to conversion, thus smaller amounts of volume—in which case the query-specific headline structure in standard ETAs were able to outperform Google’s RSAs.
Another client who couldn’t find success with RSAs cited overall low volume. The new ad type just didn’t get enough traction and barely served—making any learnings or results negligible. In the same way the new SERP inventory was beneficial for our first example, it ended up being detrimental for our last example.
If you have very limited volume, and your copy hinges on very stringent customization, or brand restrictions, RSAs might not be for you.
While adopting RSAs early gave some advertisers access to additional inventory volume, now that standard ads have been updated to match, all text ads are now serving in the same auction—so increased impression share is no longer a benefit you should expect.
Conclusion: Give It a Try
In sum: RSAs are still pretty new. Google continues to encourage automation/machine learning strategies, and will likely continue to do so. In the meantime, it’s always good to try something new to see if it works for you! So if you haven’t implemented RSAs yet, I suggest you do— but monitor them closely until you know how they perform for your accounts.