Creating better ads in a world with only RSAs

Frederick Vallaeys spent 10 years working at Google and was instrumental in helping build features like conversion tracking and analytics. For the last 10 years, he’s been running his current company Optmyzr, which is a PPC management software that helps advertisers optimize their campaigns. In his session at this year’s SMX Advanced, Vallaeys goes into detail on how to make the most out of your RSAs and automated bidding strategies.

The numbers. About 92% of advertisers have shifted to using RSAs, and the majority of those who have turned them on or used them, leave them on without shutting them off. Vallaeys hypothesizes that even years ago when advertisers didn’t know that RSAs were going to be the only option, performance was good enough that advertisers were willing to stick with them.

Vallaeys goes on to say that of the 1.7 million ads they audited, the CTR was almost exact between RSAs and ETAs. However, the conversion rate was about 11% worse with RSAs. They found that ad groups with RSAs received 1.6 times as many conversions as ad groups without RSAs. And that’s all we really want, right? More conversions.

Making the leap to RSAs. Google recommends taking your existing ETAs and start building your RSAs. Google claims that without making any changes to the ad text, creating an RSA will result in about a 7% increase in conversions with a similar cost per acquisition.

Vallaeys and the team at Optmyzr even wrote a script that you can download to make the transition from ETAs to RSAs much easier. You can download the script here.

Smart bidding strategies. Vallaeys is a fan of using smart bidding with RSAs, though when it comes to using broad match keywords, he says that’s up to the advertiser. Google did studies where they found a 20% “lift” when enabling smart bidding and broad match keywords. Though they didn’t clarify if the “lift” was an increase in clicks, CTR, or conversions.

Vallaeys reminds us that with an automated bidding strategy with RSAs, Google will show ads that match your keywords, even if the queries aren’t exactly the same. If the keyword or phrase is similar and your ad shows, you may be stuck paying a higher CPC than you normally would if the bidding were left up to Google. The chart below outlines the differences in what you could pay if your campaigns were left on Manual CPC.

Google loves scores. Ad strength is another score in your account. Vallaeys explains that ad strength scores tend to help newer advertisers, but seasoned ad veterans know that it’s not related to actual performance. Ad strength scores are suggestions gathered from other advertisers and “what everyone else does.” Limited Eligibility status is now gone and “Poor” ad strength is not an indicator of how well your ads will serve, so take it with a grain of salt.

One interesting experiment that Vallaeys outlines is that Google suggested adding specific text into the ads and indicated that it would increase the ad strength. When that text was manually added by typing the words in, the ad strength didn’t get better. But when they click the link that Google gives you to add the exact same text, the ad strength got better. This experiment may indicate that ad strength as a statistic or measuring tool isn’t that sophisticated.

RSA structure. Based on the 1.7 million ads that Optmyzr has analyzed, Vallaeys provides several tips to help make the most of our ads:

• Having two RSAs per ad group seems to be ideal for improving conversion rates
• Use labels to connect RSAs with automation
• Use pinning techniques to fake ETAs or leave ad combinations exactly where you want them (this is suggested if you have been testing for a long time and know what works)
• More headline variants lead to more impressions
• Adding DKIs and ad customizers may increase impressions, but likely decreases conversion rates
• RSAs need to be tested against RSAs. RSAs cannot be tested against ETAs simply because they’re too different
• Don’t fixate on old metrics
• Use ad variations to test and iterate creative (on the Campaign level, under the Experiments link)

Don’t miss Frederick Vallaeys session in SMX Advanced 2022

For more details on testing and optimizing RSAs and the results of the experiments Optmyzr ran, get your free pass here.

Already registered for SMX Advanced? Log in.

About The Author

Nicole Farley is an editor for Search Engine Land covering all things PPC. In addition to being a Marine Corps veteran, she has an extensive background in digital marketing, an MBA and a penchant for true crime, podcasts, travel, and snacks.