Your ads have "Good" ratings in Google Ads. Your agency says the RSA performance is strong. But your CTR hasn't improved in three months, and you still have no idea which headlines are actually driving conversions.
RSA performance labels tell you very little. What your agency does with the asset-level data tells you everything.
Why RSA Reporting Misleads Most Marketers?
Responsive Search Ads allow you to input up to 15 headlines and 4 descriptions. Google's ML combines them dynamically to show the most relevant combination for each searcher. The interface gives each asset a performance label — "Low," "Good," or "Best" — based on how often Google chooses to serve it.
But "Best" means Google shows that asset frequently. It doesn't mean that asset is responsible for conversions. An asset gets served more when it appears in many combinations that resonate with different queries — not necessarily when it's driving the specific outcomes you care about.
The black box is real. Without deliberate asset management, agencies put 15 headlines in an RSA, let Google optimize, and report a "Good" or "Excellent" overall rating without ever understanding what messaging principle is performing and why.
RSA performance labels measure Google's confidence in an asset. Your agency should be measuring what that asset tells you about your ICP's motivations.
What Expert RSA Asset Management Looks Like?
Load Truly Diverse Headlines
The most common RSA mistake is filling 15 headline slots with variations of the same message: "Award-Winning Project Management Software," "Top-Rated Project Management Tool," "Best Project Management Platform." These aren't 15 diverse assets — they're one asset with synonym substitution -- and a adwords agency builds this into every engagement.
Genuinely diverse RSAs cover different value propositions, different audiences, different CTAs, and different proof points:
Feature-led: "Real-Time Collaboration for Remote Teams"
Benefit-led: "Cut Project Review Time by 40%"
Social proof: "Trusted by 10,000+ Marketing Teams"
Problem-framed: "Stop Missing Project Deadlines"
CTA-led: "Start Your Free Trial Today"
A paid search agency building RSAs correctly ensures each headline communicates something distinct that Google can test against a different audience segment or query type.
Use Pinning Strategically, Not Universally
Pinning locks a specific headline to position 1, 2, or 3. It's useful for brand control — ensuring your brand name always appears, or that a legally required disclaimer always shows — but overused pinning defeats the purpose of RSA combination testing.
If every position is pinned, you've essentially created a standard text ad with extra steps. Pin only what must appear for legal or brand reasons. Leave the remaining positions unpinned for Google's ML to optimize.
Retire and Replace Low-Performing Assets Systematically
Assets marked "Low" by Google for more than 30 days should be retired and replaced with new variants. This is a monthly hygiene task that most agencies skip.
When you retire a "Low" asset, document why you think it underperformed. Was it the messaging angle? The specificity? The CTA framing? This analysis builds the institutional knowledge base about what resonates with your ICP.
Test One Variable at a Time Through Deliberate Experiments
True RSA testing requires more discipline than just adding headlines and watching labels. Use Google Ads experiments to A/B test specific headline concepts — run a 50/50 traffic split between RSAs with different headline sets, isolating the variable you're testing.
This is how you generate transferable learnings: "Benefit-led headlines outperform feature-led headlines for this ICP" is a conclusion you can apply across campaigns, to landing pages, and to email subject lines.
The RSA Framework to Request From Your AdWords Agency
Ask for the following every 30 days:
Asset performance report — which headlines and descriptions are labeled "Best," "Good," and "Low"
Retired assets and replacements — what was removed and what the replacement hypothesis is
Test conclusions — any A/B experiments run and what was learned
Asset diversity audit — confirmation that the 15 headlines cover multiple distinct value propositions
If your agency can't provide items 2, 3, and 4, they're not running RSAs — they're letting Google run them on autopilot.
Frequently Asked Questions
What does effective RSA asset management look like for an AdWords agency?
Effective RSA management loads 15 genuinely diverse headlines covering distinct value propositions — feature-led, benefit-led, social proof, problem-framed, and CTA-led — rather than 15 synonyms of the same message. Monthly hygiene retires assets labeled "Low" for 30+ days with documented hypotheses for why they underperformed, and Google Ads experiments run 50/50 traffic splits to generate transferable learnings about what resonates with your ICP.
What does a "Best" RSA performance label actually mean?
"Best" means Google serves that asset frequently across many combinations — it does not mean the asset drives conversions. An asset gets served more when it appears in combinations that resonate with different queries, not specifically when it influences the outcomes you care about. Without deliberate testing discipline, agencies put 15 headlines in an RSA, let Google optimize, and report an "Excellent" rating without understanding what messaging principle is performing.
How should pinning be used in Responsive Search Ads?
Pin only what must appear for legal or brand reasons — ensuring your brand name shows in position 1, for example. Overused pinning defeats the purpose of RSA combination testing; if every position is pinned you've created a standard text ad with extra steps. Leave remaining positions unpinned so Google's ML can test different combinations and you can observe which asset clusters drive better performance.
What monthly RSA report should I request from my AdWords agency?
Request four items monthly: an asset performance report showing "Best," "Good," and "Low" labels; a list of retired assets with replacement hypotheses; conclusions from any A/B experiments run; and an asset diversity audit confirming that headlines cover multiple distinct value propositions. If your agency cannot provide the last three items, they are letting Google run your RSAs on autopilot rather than actively managing them.
The Compounding Advantage of Systematic Asset Testing
Accounts that run systematic RSA testing for 6-12 months develop something valuable: a messaging library with evidence. They know that social proof headlines perform better than feature headlines in competitive categories. They know that CTA language affects CTR but not conversion rate. They know which benefit resonates most with which audience segment.
That's institutional knowledge. It makes every future campaign better — including landing pages, email sequences, and paid social creative. The agency that builds this library for your account is delivering value far beyond the platform metrics visible in the weekly dashboard.
RSA testing is not a fire-and-forget activity. It's a discipline that compounds over time. Demand that discipline from the team managing your account.
Comments
Log in or sign up to join the conversation.