It’s been three months since Google changed the world as we know it (or maybe just our world) by announcing a new wave of changes to its advertising platform. These changes included a new paid search ad format called responsive search ads (RSAs), and here at MAKA Digital we were excited to dig in and try them out, as they were built to allow for bigger headlines and descriptions than traditional ads, as well as an automated way of testing which ad text is the most successful.
Tell Me More.
There was a lot that initially seemed worthwhile about RSAs. Larger headlines and descriptions are never a bad thing for enticing viewers to click on your ads—they provide more information about your brand and they take up more real estate on the search results page. These ads also allow you to give Google a group of different headline and text ads, and Google will dynamically serve these in different combinations until it finds the combination that leads to the best click-throughs and other metrics. Google said that in early tests RSAs had led to higher click-through rates, and all these factors combined seemed to make them a slam dunk.
What’s the Catch?
When we first launched RSAs in isolated campaigns, the outlook looked good—during those first few weeks, they showed higher click-through rates and lower CPCs than traditional ads. Fast forward to several months later, though, and the prognosis doesn’t sound as great. Since those first tests we’ve run RSAs across several client accounts, and the results have changed (read: performance staled in a big hurry). Click-through rates are still occasionally higher, but often only within campaigns focusing on brand-related keywords, and sometimes even then they’re not any higher than traditional ads. And they haven’t led to a significant change in conversion rates, either. This a phenomenon that’s been echoed industry-wide.
So What Happened?
This turn toward low performance seems counterintuitive. Why wouldn’t bigger ads/longer text lead to more clicks? We’re not sure, but it seems it has something to do with Google dynamically optimizing for the best-performing text combinations—this appears to have caused click-throughs to worsen, not improve, indicating that Google’s bots are simply not as adept as human marketers in judging which text is truly the better performer. Google also has not offered much help in analyzing its selected combinations of text within an ad, with only impression data available to view here rather than click-throughs or conversions.
So Where Do We Go from Here?
At this point, we’ve taken to analyzing RSAs at an individual level—if one appears to be performing strongly for a certain group of keywords, we’ll keep it in place, and if it’s underperforming in another group of keywords, we’ll turn it off in favor of traditional ads. Though overall trends show that RSAs are not all they cracked up to be, every keyword group is different, and it’s worth weeding through for any black sheep. If you’re ready to turn off all RSAs, though, don’t be afraid—Google has promised that they won’t be sunsetting traditional ads, so they’ll always be a viable option to return to. Even better, traditional ads now offer longer headlines and descriptions, which is a great option to try if you’re not seeing success from RSAs.
In upcoming months we’ll be waiting for Google to fine-tune its dynamic text combinations and provide more metrics to compare these combinations, and maybe then we’ll give RSAs more weight. In the meantime, we’re appreciative that human marketing instincts have won at least one battle against Google’s all-powerful bots.
Still have more questions about RSAs vs. traditional ads? Did all the acronyms and metrics we mentioned sound like mumbo-jumbo? Give us a shout out at firstname.lastname@example.org and we’d be happy to talk it through.