Responsive Search Ads
Google's only Search ad format, where you provide up to 15 headlines and 4 descriptions and Google's machine learning assembles the best-performing combination for each query and user.
Your search ad copy doesn’t run as the sentence you wrote. It runs as the combination Google selected from the pool you supplied. That selection happens at auction time, matched to the person’s query, device, and prior context.
Google can assemble more than 40,000 combinations from a single RSA’s inputs.
Expanded Text Ads retired in June 2022. Call-only ads followed in February 2026. RSAs are the only text ad format remaining in Google Search campaigns.
How it shows up in the wild
Danone SGM (FMCG, Southeast Asia): The brand ran two distinct audience segments — new moms and moms-to-be — against the same RSA, letting Google assemble the most query-relevant headline combinations for each group without separate campaigns. CTR ran 8% higher than previous campaigns, per Think with Google APAC. Moms and moms-to-be visiting the site increased by 70%.
MyFlightSearch (online travel agency): The brand focused on building RSAs with “Good” or “Excellent” Ad Strength — Google’s asset-level quality score measuring headline relevance, quantity, and diversity. 14% more conversions and a 15% lower cost per booking resulted, per Google’s best practices documentation. Ad Strength was the only lever changed.
Swoop (Canadian ultra-low-cost airline): The brand ran keywords inside RSA headline assets themselves, treating asset-level keyword inclusion and campaign-level keyword targeting as two separate signals. Revenue increased 71% and conversions climbed 61%, documented in the same Google best practices source.
Why it matters
Accounts that move RSA Ad Strength from “Poor” to “Excellent” see 15% more clicks and conversions on average, per Google’s platform data. Adding a second RSA to an ad group yields a further 6.6% lift in conversions at similar cost.
The quality of your RSA inputs is now a Search performance variable in the same way creative quality is a Meta performance variable. My hunch is that most D2C Google Search accounts underinvest in RSA asset depth because the format feels automated after initial setup — the automation obscures how much the pool quality still shapes the output.
Related terms
Frequently asked questions
Should I pin headlines in RSAs?
Pin only for brand name or a compliance-required disclosure. Pinning every headline forces a static ad and removes Google’s ability to test combination performance. It defeats the format’s purpose.
Does Ad Strength predict conversion performance?
Ad Strength measures input quality — asset diversity and relevance — not conversion outcomes. An RSA rated “Average” can outperform one rated “Excellent” if the account has strong Smart Bidding signal. Use it as a diagnostic for whether you’ve given Google enough combinations to work with, not as a performance proxy.
How many RSAs should I run per ad group?
Two is the documented baseline. A second RSA yields 6.6% more conversions on average; a third adds another 3.7%. Run two with distinct messaging angles — benefit-focused and objection-focused — and let performance data separate them.