Google has been pushing the adoption of Responsive Search Ads (RSAs) for quite some time, but many advertisers have been hesitant to give Google’s algorithm control over their messaging. RSAs allow you to enter multiple headlines and descriptions, and then Google Ads will automatically test different combinations and use machine learning to determine which combinations perform best. Google prefers automated forms of bidding and creatives, as it gives them more control over spend and traffic. However, there has been some confusion on how to best measure the effectiveness of RSAs which has only increased hesitation over testing the newer ad format. Outlined below is how to write, test, and effectively measure the success of RSAs.

  • Write non-repetitive, seamless ad copy
  • Start small, measure, then scale accordingly
    • Utilize experiments to directly A/B test ad group performance
    • Measure performance of ad groups as a whole, to understand the value of the incremental traffic and conversions driven by RSAs

Check out more about (un)Common Logic’s digital marketing approach and company values here.

Read on for more information and insights from an experienced practitioner regarding how to write, test, and measure the success of Responsive Search Ads.

How to Write RSAs

When writing RSAs, the main objective will be to present a variety of enticing, non-repetitive value propositions. Non-repetitive means any combination of headlines and description lines will read seamlessly and not feel like you are serving the same information with different wording. For example, “24/7 Customer Support” and “Outstanding Customer Service” would read as repetitive if placed in the same ad by Google. If there is an important value proposition that you want to ensure is visible, you can pin the headline or description line containing that messaging by selecting the position of a specific headline as 1,2, or 3 in Google Ads or Google Ads Editor. Google will then let the algorithm do the rest of the work – using machine learning to determine which combination of value propositions is ideal for a specific user.

How to Test RSAs

When testing RSAs, a gradual rollout is optimal. Start with a couple of high traffic ad groups and create an experiment. This way you can evenly split traffic between two versions of the ad groups – one with an RSA rotated in with the existing Expanded Text Ads (ETAs), and one running only the ETAs. The A/B testing method is an effective way to prove the value of RSAs, the most important of which is additional traffic (when using RSAs, Google decides you are eligible to enter more auctions than with just ETAs). By setting up a measured A/B test, you can quickly get results that prove RSAs and their additional traffic are either worth it or not.

Are RSAs Worth It?

Now to measure the success of your test. RSA and ETA template performance should not be directly compared to each other as the ad types typically vary in performance, and RSAs will take more time to build up historical data and show improvements. Instead, you will want to look at the overall ad group performance – with and without RSAs. This is where you will see the added value of additional traffic, and consequently additional conversions. In the example below, if you look only at the RSA vs. the ETA, the ETA is clearly the better performing template in terms of conversion rate. However, when you view the data from the overall ad group level, the RSAs did aid in driving additional traffic & conversions, due to the additional specific auctions Google has inserted these ads into. It’s important to note that not all accounts will see a lower CPC with RSAs as in the example, which is why it is recommended to test into this ad format alongside the standard ETAs you already have running.

While RSAs are not always the best option for some advertisers, it is important to know how to test into this format and best measure its performance. Using the above technique will give you a better idea of the benefits of RSAs – the additional traffic and conversions are often of value.

Contact us to learn more about (un)Common Logic’s digital marketing approach.