Google announced today that, starting June 30, 2022, responsive search ads (RSAs) will be the only Search ad type that can be created or edited in standard text ad campaigns. Existing expanded text ads (ETAs) will still be eligible to serve after that date but cannot be created or updated.

This announcement doesn’t exactly come as a surprise, especially after Google made RSAs the default ad type in February. We already recognized then that Google was increasingly leaning into automation and machine learning for its ad platforms with capabilities like dynamic search ads, Smart Shopping, and Smart Bidding. Those resistant to automated solutions can continue to run ETAs created before the June 30 cut-off date. However, recent data suggests that advertisers should embrace the move to a more machine-learning based approach.

Performance for RSAs is Promising

From what we’ve observed, most advertisers should see performance benefits from a shift to RSAs. Google’s data suggests that advertisers that add RSAs in their ad groups receive up to 10% more clicks and conversions. Merkle data shows further performance benefits for activating RSAs - since Q2 2020, RSA click-through rate (CTR) has consistently been above ETAs, with accounts seeing a 12% lift on average for Q2 2021. Q2 2021 also marked the first time that RSA click share surpassed that of ETAs for Merkle clients, so the CTR benefits are holding true even as the format continues to scale. The observed CTR improvements over ETAs suggests that Google’s machine learning is doing a good job finding the right custom combination of headlines and descriptions to serve to each customer.

Best Practices for an RSA-Focused Future

Each brand’s immediate next steps will likely depend on their current RSA maturity level. Eventually, all advertisers should adhere to the following best practices to maximize the opportunity afforded by RSAs.

• Brands without at least one RSA per ad group should focus on gaining full coverage. While ETAs will still be eligible to serve, we predict that RSAs will be favored due to their greater relevance and performance potential, making it important to have full coverage of RSAs across ad groups. If you’re looking for a place to start, consider pulling assets from your existing ETAs or using high-performing keywords in new copy variations.
• Give Google flexibility to optimize. When launching an RSA, try to fill all 15 headline and 4 description slots to give Google as much flexibility as possible to test different combinations for different users. Note that within RSAs there’s an option to pin an asset to a certain position. There are certainly use cases for this – keeping a branded headline visible in position 1, for example, or pinning creative within verticals that have strict copy requirements. However, where possible, it’s best to use pinning sparingly to give Google’s algorithm the chance to experiment, learn, and optimize.
• Monitor ad strength. As with other ad formats, Google provides ongoing feedback to help advertisers improve the quality of their RSAs. Checking ad strength regularly and making updates to improve it helps advertisers better position themselves for success with the format. To point advertisers in the right direction, Google calls out specific areas for improvement like having headlines that are too similar or not enough keywords in the descriptions.
• Target to the ad group. As with all ad copy, messaging should be hyper-relevant to potential searchers based on the keywords contained in the ad group. Having 19 different assets to rearrange and swap is good, but having 19 different highly relevant assets is even better. One way to do that is to…
• Incorporate dynamic elements where applicable. RSAs can now handle countdown and location-based customizers, providing additional opportunities to make ads more relevant.
• Refresh on a recurring basis. Ensure that copy is reviewed and updated regularly to account for seasonality as well as changing consumer behaviors and sentiments. Look for individual assets that aren’t getting much visibility and consider replacing them to give Google new variations to test. One caveat – unfortunately, only impression data is currently available at the asset and combination level. We hope to eventually see more insightful metrics, like CTR, to better understand how each individual asset performs.

Overall, given the current data on RSAs, we’re excited to see Google streamline its ad copy options. The push towards a more automated, machine-learning based format should prove beneficial for advertisers and customers alike. Meanwhile, advertisers that want to keep full control over their messaging can still do so by using the pin feature or creating the evergreen ETAs they want to use prior to June 30, 2022 – though we’d highly recommend giving RSAs a chance!

It’s great to see new campaign types that cater specifically for video content. What I’m especially excited about here is that advertisers will be able to create lookalike segments from first-party and YouTube audiences, that have customisable reach. Want to keep things tight? A narrow segment will hit only 2.5% of users with traits like that in your seed list. Want to scale? Go for 10% with their broad reach option. Great for taking control in a world where it seems to slip away from us at times.

It will be a truly exciting time for us as these products start to launch over the next few months. If you’re interested in learning more about what these developments could mean for you, get in touch and let’s make the most of AI together.