When it comes to assessing paid media performance, many advertisers are comparing Responsive Search Ads (RSAs) with Expanded Text Ads (ETAs). However, this isn’t a fair comparison as Google has announced that RSAs are being shown on inventory where ETAs wouldn’t be eligible to appear. For example, RSAs can be shown on searches where ETAs may have had a low Ad Rank. As a result, RSAs will enable advertisers to broaden their reach and, as a rule of thumb, should be evaluated based on incrementality.

Why RSAs help advertisers increase reach

• As RSAs have a variety of components to build an ad copy, the auction will try to find a combination that will deliver a high-quality score and qualify for searches. In contrast, ETAs might have a low-quality score and not be eligible to serve the ad.
• Machine learning helps create an ad that’s relevant enough to access inventory ETAs weren’t able to access.

The new inventory RSAs can be found by running a search term report. The best way to review the specific new placements is by comparing the period right before and after RSAs were enabled in an ad group. From there, new queries where there were no impressions before but that led to an increase in impressions after the RSAs were added can be identified.

Help the machine learn

It’s essential to review the inventory your RSAs are served on long after you first implemented RSAs, as these ads use machine learning. Through machine learning, the system learns which inventory works poorly and which is promising. Google will automatically turn off inventory that’s performing poorly and push the best performing ones. That’s why it’s worth reviewing the inventory on a regular basis by looking at ad performance segmented by query, using a script. The script will allow you to review 46 possible reports and will help you automate the process of downloading each report separately. This way, more informed decisions can be made about the RSAs and ad components can be refreshed.

RSAs performance and reach can only be as good as the ad components advertisers provide. Therefore, it’s essential to give machine learning time to learn and for the advertiser to review its impact before experimenting with other ad components. Google has guidelines to help create the best possible components.

Many advertisers have been struggling to see the true value of RSAs, but now that more information on how to review ads is available, they can now make more informed decisions moving forward.