Responsive Search Ads (RSAs) have become the standard ad format within Google Ads search campaigns. Since expanded text ads were phased out, advertisers now rely on RSAs as the primary way to deliver messaging within search results.
However, one question frequently arises when structuring campaigns:
How many Responsive Search Ads should you have in each ad group?
For many advertisers who were accustomed to running multiple expanded text ads for testing, the shift to RSAs changed how ad testing works. Instead of comparing separate ads, Google now tests different combinations of headlines and descriptions within a single RSA.
This shift has led to some confusion about whether advertisers should run one RSA per ad group, multiple RSAs, or some combination of both.
The answer depends on how RSAs function and what role they play within Google’s automated system.
This article explores how Responsive Search Ads best practices and how they work within ad groups, what Google recommends and what advertisers should consider when deciding how many RSAs to run.
A Quick Refresher: How Responsive Search Ads Work
Before discussing how many ads should exist within an ad group, it is important to understand how Responsive Search Ads operate.
When creating an RSA, advertisers can provide up to:
15 headlines
4 descriptions
Google’s system then automatically tests different combinations of these assets to determine which versions perform best.
In most search results, the ad will display up to three headlines and two descriptions. However, the exact layout may vary depending on the device and available space.
Each time the ad enters an auction, Google evaluates contextual signals such as:
The user’s search query
Location
Device type
Time of day
User behaviour patterns
Based on these signals, the system selects the combination of headlines and descriptions most likely to perform well.
Over time, machine learning models analyse engagement and conversion data to prioritise the combinations that deliver the strongest results.
Because of this dynamic testing process, a single Responsive Search Ad can generate hundreds or even thousands of potential ad combinations.
This is a key reason why the role of multiple ads within an ad group has changed.
How Ad Testing Used to Work
Before Responsive Search Ads were introduced, advertisers typically created several expanded text ads (ETAs) within each ad group.
These ads would rotate against one another, allowing advertisers to compare performance between different versions.
A typical structure might have included:
Three or four ads within each ad group
Different headlines and descriptions in each ad
Manual comparison of click-through rates and conversion rates
Over time, underperforming ads would be paused and replaced with new variations.
This process was known as A/B testing or multivariate testing.
However, Responsive Search Ads changed this approach.
Instead of advertisers manually testing different ads, Google now tests combinations automatically within a single RSA.
This raises the question of whether running multiple RSAs is still necessary.
Google’s Recommendation
Google generally recommends running at least one Responsive Search Ad per ad group.
In many cases, a single well-constructed RSA provides enough variation for the system to test and optimise effectively.
Because a single RSA can contain up to fifteen headlines and four descriptions, it already offers a wide range of potential combinations.
If the assets are diverse and thoughtfully written, Google’s system can test many variations without requiring multiple ads.
Google often encourages advertisers to focus on improving asset diversity within a single RSA rather than creating several similar ads.
This approach aligns with the platform’s broader shift toward automation.
Instead of comparing separate ads, Google’s system determines which combinations of messaging perform best.
Why One RSA Is Often Enough
In many ad groups, a single Responsive Search Ad is sufficient.
This is particularly true when the RSA contains a strong set of assets that cover multiple value propositions and keyword themes.
For example, an RSA might include headlines focused on:
Pricing
Experience
Speed of service
Customer satisfaction
Industry specialisation
Because Google can mix and match these elements, the system can test many different variations.
In practice, a single RSA with fifteen headlines and four descriptions can generate thousands of potential ad combinations.
This level of variation often provides more testing opportunities than several fixed ads ever could.
For this reason, many advertisers now run one primary RSA per ad group.
When Multiple RSAs Might Make Sense
Although one RSA is often sufficient, there are situations where running multiple RSAs can still be useful.
One example involves testing different messaging strategies.
Instead of placing all messaging variations within a single RSA, advertisers might create separate RSAs that emphasise different themes.
For example:
One RSA focusing on price and affordability
Another focusing on expertise and experience
This allows advertisers to test broader messaging approaches rather than simply testing small headline variations.
However, when using multiple RSAs, it is important to ensure that each ad has distinct messaging.
Creating several RSAs with nearly identical headlines and descriptions rarely adds value.
Instead, it can dilute the data available to each ad and slow down optimisation.
Potential Drawbacks of Running Multiple RSAs
While running multiple RSAs may seem like a good way to test additional variations, it also introduces certain challenges.
One issue is data fragmentation.
If multiple RSAs exist within the same ad group, impressions and clicks are divided between them.
This means it may take longer for each ad to accumulate enough data for meaningful optimisation.
In contrast, concentrating impressions within a single RSA allows the system to learn more quickly which combinations perform best.
Another issue is asset overlap.
If several RSAs contain similar headlines and descriptions, the system may effectively be testing the same messaging repeatedly.
This can reduce the effectiveness of the testing process.
Because of these factors, many advertisers prefer to focus on building one strong RSA rather than multiple similar ones.
The Role of Asset Diversity
Regardless of how many RSAs are used, asset diversity remains one of the most important factors influencing performance.
Google’s system works best when it has access to a wide range of messaging options.
This means headlines should highlight different benefits rather than repeating similar phrases.
For example, weak headline variation might look like this:
“Professional Accounting Services”
“Reliable Accounting Services”
“Expert Accounting Services”
A stronger set of headlines might include:
“Specialist Small Business Accountants”
“Reduce Your Business Tax Liability”
“Over 20 Years of Accounting Experience”
“Fixed Fee Accounting Packages”
These variations allow Google to test different value propositions and determine which resonate most strongly with users.
When asset diversity is strong, a single RSA can often outperform multiple weaker ads.
What About Ad Strength?
Google Ads provides a metric known as Ad Strength, which evaluates the diversity and relevance of assets within a Responsive Search Ad.
Ads with more varied headlines and descriptions typically receive higher Ad Strength ratings.
While Ad Strength should not be treated as a direct performance metric, it does provide useful guidance.
If an RSA receives a low Ad Strength rating, it may indicate that additional or more varied assets are needed.
Improving asset diversity can strengthen the ad’s ability to adapt to different search queries.
This often reduces the need for multiple separate RSAs.
How Agencies Typically Structure Ad Groups
Many experienced Google Ads practitioners now adopt a relatively simple structure.
A common approach includes:
One primary Responsive Search Ad per ad group
Strong asset diversity within that ad
Minimal pinning to allow flexibility
In some cases, a second RSA may be added to test a different messaging approach.
However, the focus usually remains on providing high-quality assets rather than increasing the number of ads.
This approach aligns with Google’s automated testing model.
By concentrating impressions within a smaller number of ads, the system can gather performance data more quickly.
How to Test Messaging Effectively
Advertisers who want to test different messaging strategies should think carefully about how they structure their RSAs.
One approach is to include multiple value propositions within a single RSA and allow Google to test combinations automatically.
Another approach involves creating separate RSAs with distinct messaging themes.
For example:
RSA A: Focused on pricing and affordability
RSA B: Focused on expertise and reputation
Over time, advertisers can evaluate performance metrics such as click-through rate and conversion rate to see which messaging resonates most strongly.
However, it is important to ensure that each RSA has sufficient impressions to generate meaningful data.
Running too many ads within a single ad group can make testing less reliable.
Final Thoughts
Responsive Search Ads fundamentally changed how ad testing works within Google Ads.
Instead of manually testing multiple fixed ads, advertisers now provide a range of assets that Google’s system automatically combines and optimises.
Because of this shift, the number of ads within an ad group is less important than the quality and diversity of the assets within those ads.
In many cases, a single well-constructed RSA is enough to generate a wide range of ad variations and allow Google’s system to optimise effectively.
Some advertisers may choose to run multiple RSAs when testing different messaging strategies, but this should be done carefully to avoid fragmenting data.
Ultimately, the most important factor is ensuring that Responsive Search Ads include strong, varied messaging that reflects user intent.
When this foundation is in place, Google’s machine learning systems are better able to identify the combinations that drive the strongest results.