Responsive Search Ads for SaaS: How to Actually Control What Google Shows
Sabarish Chandrasekar
Last updated on
Responsive Search Ads (RSAs) are Google’s ad format that combines up to 15 headlines and 4 descriptions you provide, testing different combinations automatically to find what performs best for each auction. They’re now the default and only standard search ad format in Google Ads, which makes structuring assets deliberately, not just writing them, the actual skill.
What You Need to Know About Responsive Search Ads
- RSAs combine your submitted headlines and descriptions automatically. You don’t control the exact final combination shown to any given searcher.
- Google’s system tests combinations against each other, favoring ones that perform better over time, similar to a built-in, continuous A/B test.
- Pinning a headline or description locks it to a specific position, trading some optimization flexibility for message control, and it’s underused by most advertisers who could benefit from it.
- Ad Strength (Google’s asset-quality score, shown as Poor, Average, Good, or Excellent) correlates with performance but isn’t a guarantee, and chasing “Excellent” for its own sake can produce generic, diluted messaging.
- RSAs replaced Expanded Text Ads, which Google fully deprecated. There’s no manual, fixed-combination alternative left for standard search ads.
What Are Responsive Search Ads?
Responsive Search Ads let you submit a pool of up to 15 headlines and 4 descriptions, then Google’s system dynamically assembles and tests different combinations of those assets for each individual auction, learning over time which combinations perform best for which searches.
This is a fundamentally different model from the old Expanded Text Ads format, where you wrote one fixed headline-and-description combination per ad. With RSAs, you’re not writing a single ad. You’re supplying raw material for an algorithm that assembles many possible ads from it, and you rarely see the exact combination a given searcher saw.
- Up to 15 headlines, each limited to 30 characters, though you don’t need to use all 15 slots.
- Up to 4 descriptions, each limited to 90 characters.
- Google displays 2-3 headlines and 1-2 descriptions at a time, assembled in different orders and combinations depending on the auction and device.
- Pinning locks a specific asset to a specific position (headline 1, for example), which removes it from the automated testing but guarantees it always appears there.
Consider a SaaS company that submitted 15 headlines without pinning any of them, including several with contradictory value propositions (one emphasizing “enterprise-grade security,” another emphasizing “the fastest setup in the industry”). Google’s system occasionally combined the two in ways that read as inconsistent to searchers, since nothing constrained which pairs could appear together.
How to Structure RSA Assets So Google Doesn’t Scramble Your Message
The core skill in Responsive Search Ads isn’t writing 15 good headlines in isolation. It’s structuring the pool so that any combination Google assembles still reads as coherent, since you don’t control which ones pair together.
How to Build a Responsive Search Ad Step by Step
- Write headlines in thematic groups, rather than as 15 unrelated ideas. Group headlines around 2-3 core themes (a specific benefit, a proof point, a call to action) so combinations stay coherent regardless of which ones Google pairs.
- Pin your most essential message to a fixed position. If a specific claim, like a certification or a specific price point, must always appear, pin it rather than leaving it to chance.
- Avoid pinning everything. Over-pinning defeats the purpose of RSAs by removing Google’s ability to test and optimize, functionally turning the format back into a fixed ad with extra steps.
- Include keyword-relevant headlines without keyword-stuffing every slot. A few headlines that closely match likely search terms help relevance, but filling all 15 with keyword variations produces a repetitive, low-quality-feeling ad, similar to how Performance Max rewards diverse, non-redundant creative assets.
- Write descriptions that can stand alone or pair with any headline. Since you don’t control which headline-description pairs appear together, each description needs to make sense regardless of which headline precedes it.
- Check Ad Strength, but don’t chase “Excellent” at the cost of message clarity. A “Good” rating with sharp, differentiated messaging usually outperforms an “Excellent” rating built from generic, keyword-diluted filler.
- Test multiple RSAs per ad group with genuinely different angles. Running 2-3 RSAs with distinct core angles (cost savings, speed, security) gives you a real signal about which value proposition resonates, which a single RSA can’t isolate on its own.
Common Mistakes to Avoid
Writing 15 Headlines Without a Coherence Check
If any two headlines could appear together and contradict or clash with each other, that combination will eventually show to a real searcher, since Google tests broadly across the full asset pool.
Chasing “Excellent” Ad Strength Over Message Quality
Ad Strength measures asset diversity and keyword relevance, not persuasiveness. A technically “Excellent” ad built from generic, repetitive headlines can underperform a “Good” ad with sharper, more differentiated copy, the same quality score tradeoff that shows up across the rest of the account.
Pinning Every Asset
Pinning every headline and description to a fixed position removes Google’s ability to test combinations, which defeats the purpose of the format and often performs worse than a properly structured, partially pinned RSA.
Running Only One RSA Per Ad Group
A single RSA, however well-built, only tests combinations within its own asset pool. Running multiple RSAs with genuinely different core angles gives you a real read on which message resonates, not just which combination of similar messages performs marginally better. Our Google Ads benchmarks show how much this discipline moves conversion rate across real SaaS accounts.
How PipeRocket Digital Builds Responsive Search Ads
We build RSAs in thematic asset groups with strategic pinning, then run multiple distinct RSAs per ad group to test genuinely different angles, rather than treating asset creation as filling 15 blank slots. For the copywriting approach behind the assets themselves, see our guide on writing Google Ads copy for SaaS , or our SaaS PPC team can build the full account structure around it. Get in touch if your RSAs are underperforming and you’re not sure why.
Frequently Asked Questions
How many headlines and descriptions should I actually submit?
Google allows up to 15 headlines and 4 descriptions, but submitting the maximum isn’t automatically better. Aim for enough variety across 2-3 clear themes to give the algorithm real combinations to test, typically 8-12 headlines and all 4 descriptions, rather than padding to 15 with redundant or weak variations that dilute Ad Strength and message clarity.
What happens if I pin a headline to position 1?
The pinned headline always appears in that position when the ad shows, and Google won’t test alternatives there. Other unpinned headlines and descriptions continue rotating in the remaining positions. This trades some optimization potential for guaranteed message control, which is worth it for claims that must always appear, like a required disclaimer or your core value proposition.
Can I still run Expanded Text Ads instead of RSAs?
No. Google fully deprecated Expanded Text Ads, and Responsive Search Ads are now the only standard search ad format available for new ad creation. Any existing Expanded Text Ads still running were grandfathered in for reporting purposes but can no longer be created or edited into new campaigns.
