TL;DR: Google's Ad Strength metric is a compliance score, not a performance score. A study of 93,055 RSAs shows it has zero impact on Ad Rank or Quality Score. What actually matters: 2 RSAs per ad group, keywords in only 20-30% of headlines, strategic pinning decisions, and a weekly testing cadence using the hidden Ad Variations feature.
Google wants you to chase "Excellent" Ad Strength. The data says you should chase conversions instead. This guide shows you how.
The RSA Landscape in 2026
Responsive Search Ads are no longer optional. With Expanded Text Ads (ETAs) sunset, 92% of advertisers now run RSAs as their primary ad format. RSAs allow you to provide up to 15 headlines and 4 descriptions, and Google's machine learning combines them into different variations to find what works.
The promise: Google tests thousands of combinations so you do not have to.
The reality: Google optimizes combinations for click volume (which generates revenue for Google), not for conversion quality (which generates revenue for you).
Understanding this structural conflict is the foundation of effective RSA optimization.
Related reading: For the full breakdown of how Google's incentives conflict with yours, see Google Ads Efficiency Playbook: The Foundation.
RSA Performance Data: What 93,055 Ads Tell Us
Optmyzr analyzed 93,055 RSAs across 13,671 accounts to understand real-world performance patterns. The findings contradict most of Google's own recommendations.
RSAs vs ETAs: The Volume Trade-Off
| Metric | RSAs | ETAs | Difference |
|---|---|---|---|
| Impressions | Baseline | 4x fewer | RSAs win on reach |
| Average conversion improvement | +7% | Baseline | Slight RSA edge |
| Advertiser adoption | 92% | Declining | RSAs are the standard |
RSAs deliver 4x the impressions of ETAs. When migrating from ETAs with identical assets, advertisers see an average 7% conversion improvement. One case study showed conversion rates improving from 9.19% to 10.2%, generating 200 additional weekly conversions.
But volume is not the whole story.
The Optimal Number of RSAs Per Ad Group
This finding surprised many advertisers:
| RSAs Per Ad Group | Result |
|---|---|
| 1 RSA | Lower conversion rates |
| 2 RSAs | Optimal conversion rate (sweet spot) |
| 3 RSAs | More impressions but diluted conversion performance |
| 4+ RSAs | Diminishing returns, data fragmentation |
Two RSAs per ad group is the conversion sweet spot. Not one, not three. Two.
Why? With a single RSA, Google has limited combinations to test. With three or more, impression volume gets split across too many ads, preventing any single RSA from accumulating enough data to optimize effectively.
The Maximum Asset Configuration
Each RSA accepts:
| Component | Available Slots | Displayed | Character Limit |
|---|---|---|---|
| Headlines | 15 | 3 | 30 characters |
| Descriptions | 4 | 2 | 90 characters |
| Final URL path fields | 2 | 2 | 15 characters each |
Use all 15 headlines and all 4 descriptions. Every empty slot is a missed testing opportunity. More assets give Google's algorithm more combinations to explore, and the data consistently shows that fuller RSAs outperform sparse ones.
Ad Strength: The Metric That Doesn't Matter
This is the most important thing to understand about RSAs in 2026:
Ad Strength does NOT impact Ad Rank.
Ad Strength does NOT impact Quality Score.
Ad Strength does NOT correlate with actual performance.
What Ad Strength Actually Measures
Ad Strength is a compliance score. It measures how well your RSA follows Google's formatting recommendations:
- Did you fill all 15 headline slots?
- Are your headlines diverse enough?
- Did you include keywords?
- Are your descriptions varied?
Each improvement in Ad Strength (from "Poor" to "Average," or "Average" to "Good") yields approximately a 3% click lift. That is the total benefit. Not a 3% conversion lift. A 3% click lift. Clicks that Google charges you for.
The Evidence
An "Excellent" Ad Strength rating means you followed Google's best practices. It does NOT mean the ad will perform better. Poor-strength ads regularly deliver exceptional conversion results because they may be tightly targeted, highly specific, and precisely matched to searcher intent.
Practical example: A law firm running an RSA with pinned headlines like "24/7 Emergency Criminal Defense Attorney" might get a "Poor" Ad Strength score because of limited headline variety. But that ad converts at 3x the rate of a "Good" strength ad with generic variations like "Legal Help When You Need It" and "Experienced Lawyers Near You."
What to Do Instead
Stop chasing Ad Strength. Focus on these actual performance drivers:
- Conversion rate within your target CPA
- Conversion volume at acceptable cost
- Click-through rate as a secondary signal
- Cost per conversion trending down over time
The Keyword Headline Rule: 20-30% Maximum
Google recommends stuffing keywords into your headlines. The data says otherwise.
The Optimal Keyword Ratio
Include keywords in only 20-30% of your headlines (3-4 out of 15).
| Headline Category | Quantity | Purpose |
|---|---|---|
| Keyword-focused | 3-4 headlines | Relevance and Quality Score |
| Benefit-focused | 4-5 headlines | Value proposition |
| CTA-focused | 2-3 headlines | Drive action |
| Trust/proof-focused | 2-3 headlines | Social proof and authority |
| Differentiator-focused | 2-3 headlines | Competitive advantage |
Why Not More Keywords?
When you put keywords in every headline, you get 15 variations of the same message. Google may show 3 keyword headlines simultaneously, creating ads like:
Emergency Plumber Houston | Houston Emergency Plumber | 24/7 Plumber Houston
That reads like spam, not a compelling ad. A better combination:
Emergency Plumber Houston | Fixed in 60 Minutes or Free | 4.9 Stars - 500+ Reviews
The second version includes one keyword headline for relevance, one benefit headline for value, and one trust headline for credibility.
Headline Category Framework
Here is a practical framework for your 15 headlines:
Category 1: Keyword Headlines (3-4)
- Include your primary keyword naturally
- Match the search query the user typed
- These drive Quality Score relevance
Category 2: Benefit Headlines (4-5)
- What does the customer get?
- Focus on outcomes, not features
- Examples: "Fixed in 60 Minutes," "Save Up to 40%"
Category 3: CTA Headlines (2-3)
- Direct action language
- Examples: "Get Your Free Quote Today," "Book Online in 30 Seconds"
Category 4: Trust/Proof Headlines (2-3)
- Social proof and authority signals
- Examples: "4.9 Stars From 500+ Reviews," "Licensed & Insured Since 2005"
Category 5: Differentiator Headlines (2-3)
- What makes you different from competitors?
- Examples: "No Hidden Fees Ever," "Same-Day Service Guaranteed"
Description Strategy
For your 4 descriptions (90 characters each):
| Description # | Focus | Example |
|---|---|---|
| 1 | Primary value proposition + CTA | "Expert [service] with a satisfaction guarantee. Call now for a free estimate." |
| 2 | Trust and credibility | "Serving [city] for 15+ years. Licensed, insured, and background-checked technicians." |
| 3 | Specific offer or differentiator | "No overtime charges. Same flat rate 24/7. Get your free quote in minutes." |
| 4 | Urgency or social proof | "500+ 5-star reviews. Book today and get priority scheduling." |
The Pinning Debate: Data vs Dogma
Pinning forces specific headlines or descriptions into fixed positions. Google discourages pinning. The data tells a more nuanced story.
The 93K RSA Pinning Study
From the Optmyzr study of 93,055 RSAs:
| Configuration | Impressions | CTR | CVR |
|---|---|---|---|
| No pinning | Highest (baseline) | Lower | Lower |
| Full pinning | 3.9x fewer | Higher | Higher |
Fully pinned ads receive 3.9x fewer impressions per ad group. But they deliver higher CTR and higher conversion rates.
This is the trade-off: volume vs. quality.
When to Pin
Pin when:
- You have legal or compliance requirements (regulated industries)
- You have proven headline winners from historical data
- Your brand name must always appear in position 1
- You are running a specific promotion with required language
- You serve a local market where location must be visible
Do not pin when:
- You are launching a new campaign without historical data
- You want maximum impression volume for data collection
- You are in the testing phase of a new service offering
The Hybrid Approach
The best strategy for most service businesses:
- Pin position 1 to your strongest keyword headline (ensures relevance)
- Pin position 2 to nothing (let Google test)
- Pin position 3 to nothing (let Google test)
- Pin description 1 to your primary CTA
- Leave description 2 unpinned
This gives you control over the most important position while allowing Google to test the remaining slots.
Multi-Pin Strategy
You can pin multiple headlines to the same position. Google will rotate among them:
Position 1 (pin 3 headlines):
- "Emergency Plumber Houston"
- "24/7 Plumbing Service Houston"
- "Houston Emergency Plumber - Fast"
This gives Google 3 relevant options for position 1 while ensuring the first thing users see is always relevant to their search.
Ad Copy Testing Framework
Traditional A/B testing is dead for RSAs. Google rotates combinations automatically, making it impossible to isolate variables within a single RSA. But that does not mean you should stop testing.
The Ad Variations Feature (Hidden Gem)
Most advertisers do not know this feature exists.
Location: Campaigns > Experiments > Ad Variations
This is the proper way to A/B test ad copy in the RSA era. It allows you to:
- Test across multiple campaigns simultaneously
- Run true 50/50 traffic splits
- Get statistical significance indicators
- Isolate single variables
Step-by-Step Setup
- Navigate to Campaigns > Experiments > Ad Variations
- Click the "+" button to create a new variation
- Select target campaigns (you can filter by headline, description, or URL)
- Choose your modification type:
| Method | Best For | How It Works |
|---|---|---|
| Find and Replace | Testing CTAs, USPs, specific words | Swaps one word/phrase across all ads |
| Update URLs | Testing landing pages | Changes destination URL |
| Update Text | Adding, removing, or pinning assets | Modifies specific headlines/descriptions |
- Name your experiment clearly (e.g., "CTA Test: Free Quote vs Book Now - March 2026")
- Set start and end dates (maximum 84 days)
- Set traffic split to 50/50 (standard)
- Launch and wait for data
What to Test (One Element at a Time)
Never test multiple variables simultaneously. You cannot determine what caused the change.
Test priority order for service businesses:
| Priority | Element | Example Test |
|---|---|---|
| 1 | CTA language | "Get a Free Quote" vs. "Book Now" vs. "Call Today" |
| 2 | Primary value prop | "24/7 Service" vs. "Same-Day Guaranteed" |
| 3 | Trust signals | "500+ Reviews" vs. "Licensed Since 2005" |
| 4 | Pricing language | "Starting at $99" vs. "Affordable Rates" |
| 5 | Urgency elements | "Limited Availability" vs. "Book for Today" |
| 6 | Landing page | Service page vs. dedicated landing page |
| 7 | Display path | /free-quote vs. /emergency-service |
Statistical Significance
Do not make decisions on insufficient data.
Minimum thresholds before concluding a test:
- At least a few hundred clicks per variation
- At least 1,000 impressions per variation
- Blue star icons in Ad Variations indicate statistical significance
- Run for minimum 2 weeks, ideally 4 weeks
Interpreting Results
Review results at three levels:
- Campaign level — Overall winner
- Ad group level — Winners may differ by theme
- Individual ad level — Granular insights
A CTA that wins for emergency plumbing ("Call Now") may lose for kitchen renovation ("Get Your Free Design Consultation"). Test conclusions are context-dependent.
The Weekly RSA Testing Cadence
Consistent testing compounds results over time. Establish a weekly rhythm.
Friday Morning Protocol (1-1.5 Hours)
Step 1: Review running experiments (15 minutes)
- Check Ad Variations for statistical significance
- Document winners and losers
- End experiments that have reached significance
Step 2: Analyze RSA asset performance (20 minutes)
- Review the Combinations Report (Ads > Assets > Combinations)
- Identify which headline/description combinations get the most impressions
- Note: High impressions does NOT equal high conversions. Check both.
Step 3: Launch new test (20 minutes)
- Based on previous learnings, set up the next experiment
- Change ONE variable from the winning version
- Name clearly with date and variable being tested
Step 4: Review ad group health (15 minutes)
- Confirm each ad group has exactly 2 RSAs
- Check that all RSAs use 15 headlines and 4 descriptions
- Identify any RSAs with "Poor" asset performance and plan replacements
Step 5: Document everything (15 minutes)
- Record: What was tested, what won, by how much
- Update your testing backlog
- Plan next week's test
The Testing Flywheel
Over time, this creates a compounding effect:
Month 1: Test CTAs across 3 campaigns. Winner: "Get Your Free Quote" (+12% CTR)
Month 2: Test value propositions with winning CTA. Winner: "Same-Day Service" (+8% CVR)
Month 3: Test trust signals with winning CTA + value prop. Winner: "500+ 5-Star Reviews" (+15% CVR)
Month 4: You now have an RSA built from 3 months of data-driven decisions. Every element has earned its place.
This methodical approach yields far better results than writing RSAs based on intuition.
RSA Optimization Checklist for Service Businesses
Initial Setup
- Set up exactly 2 RSAs per ad group
- Fill all 15 headline slots in each RSA
- Fill all 4 description slots in each RSA
- Include keywords in only 3-4 of 15 headlines (20-30%)
- Create headlines across all 5 categories (keyword, benefit, CTA, trust, differentiator)
- Ignore Ad Strength rating — focus on conversion data
- Set ad rotation to "Do not optimize" for testing control
Headline Writing Checklist
- Each headline highlights a different point (no duplicates)
- Mix of short and long headlines (use the full 30 characters where appropriate)
- At least 2 headlines include specific numbers (pricing, reviews, years, speed)
- At least 2 headlines include a clear CTA
- Location included in at least 1 headline (for local service businesses)
- No headline makes false promises or uses clickbait
Description Writing Checklist
- Each description has a unique angle (value, trust, offer, urgency)
- Primary CTA appears in at least 2 descriptions
- Specific details included (not generic claims)
- Character limits maximized (use all 90 characters)
Pinning Strategy
- Pin only with historical data supporting the decision
- Consider multi-pin (2-3 headlines pinned to position 1)
- Leave at least 2 headline positions unpinned
- Document pinning rationale for future reference
Testing Protocol
- One experiment running at all times
- Ad Variations feature used for controlled testing
- Only one variable changed per test
- Minimum 2-week test duration
- Statistical significance confirmed before concluding
- Results documented in a testing log
- Weekly review cadence established (Friday mornings recommended)
Common RSA Mistakes (and How to Fix Them)
Mistake 1: Chasing Ad Strength
The problem: You rewrite headlines to improve Ad Strength from "Good" to "Excellent," making them more generic in the process.
The fix: Ignore Ad Strength entirely. Measure success by conversion rate and cost per conversion. A "Poor" Ad Strength ad that converts at 12% beats an "Excellent" ad that converts at 3%.
Mistake 2: Keywords in Every Headline
The problem: All 15 headlines contain your primary keyword. Google shows 3 keyword-stuffed headlines together.
The fix: Keywords in 3-4 headlines only. Fill the rest with benefits, CTAs, trust signals, and differentiators.
Mistake 3: Too Many RSAs Per Ad Group
The problem: You have 4-5 RSAs competing for impressions in one ad group. None accumulates enough data to optimize.
The fix: Reduce to exactly 2 RSAs. Pause the lowest performers.
Mistake 4: Never Testing
The problem: You set up RSAs once and never touch them again.
The fix: Use the Ad Variations feature for continuous testing. One test running at all times. Weekly review cadence.
Mistake 5: Testing Multiple Variables
The problem: You change the CTA, value proposition, and landing page simultaneously. Something improved, but you do not know what.
The fix: Change ONE element per test. Document the result. Then test the next element.
Mistake 6: Letting Google Optimize Rotation
The problem: Google's "Optimize" ad rotation sends 80% of traffic to one ad, starving the other of data and preventing proper testing.
The fix: Set ad rotation to "Do not optimize" (rotate ads indefinitely) when running tests. This ensures both RSAs get equal data.
Mistake 7: Generic Descriptions
The problem: Descriptions say "We provide quality service" and "Contact us today for more information."
The fix: Get specific. "Licensed plumbers with 4.9-star rating from 500+ reviews. Same-day service, flat-rate pricing, no overtime charges. Call for your free estimate."
RSA Strategy by Business Type
Emergency Services (Plumbing, HVAC, Electrical, Locksmith)
Headline priorities:
- Urgency: "Available Right Now," "60-Minute Response"
- Trust: "Licensed & Insured," "Background-Checked"
- Location: "[City] Emergency Service"
Pin position 1 to your strongest urgency headline. Emergency searchers need immediate reassurance.
Professional Services (Law, Accounting, Consulting)
Headline priorities:
- Expertise: "20+ Years Experience," "Specialized In [Practice Area]"
- Results: "97% Case Success Rate," "$10M+ Recovered"
- CTA: "Free Consultation," "No Fee Unless We Win"
Pin position 1 to your expertise or result headline. Professional service buyers need confidence signals.
Home Improvement (Remodeling, Landscaping, Painting)
Headline priorities:
- Portfolio: "See Our 500+ Projects," "Before & After Gallery"
- Value: "Free Design Consultation," "Financing Available"
- Trust: "25-Year Warranty," "A+ BBB Rating"
Leave all positions unpinned initially. Test to find what resonates with your market.
Health & Wellness (Dentists, Chiropractors, Med Spas)
Headline priorities:
- Comfort: "Gentle & Caring Approach," "Sedation Options"
- Offers: "New Patient Special: $99 Exam," "Free Whitening"
- Convenience: "Same-Day Appointments," "Weekend Hours"
Pin position 1 to your primary offer. Healthcare searchers respond strongly to specific promotions.
Measuring RSA Success
Primary Metrics
| Metric | What It Tells You | Target Direction |
|---|---|---|
| Conversion rate | Are clicks turning into leads? | Higher is better |
| Cost per conversion | Is each lead affordable? | Lower is better |
| Conversion volume | Are you getting enough leads? | Higher within budget |
| Impression share | Are you showing up enough? | 80%+ for non-brand |
Secondary Metrics
| Metric | What It Tells You | Target Direction |
|---|---|---|
| CTR | Are ads compelling enough to click? | Higher is better |
| Quality Score | Is the ad/keyword/landing page aligned? | 7+ is good |
| Ad Strength | Nothing useful | Ignore |
The Combination Report
Access via Ads > Assets > Combinations
This report shows which headline/description combinations Google serves most often. Use it to:
- Identify headlines that never appear (consider replacing them)
- Spot high-impression combinations with low conversion rates (a Google revenue problem, not a performance signal)
- Find hidden winners that appear less often but convert well
Critical insight: High impressions in the Combinations Report does NOT mean high performance. Google shows combinations that get clicks. Clicks generate revenue for Google. Always cross-reference with conversion data.
Display Path Optimization
The two display path fields (15 characters each) in your RSA are often overlooked but contribute to both relevance and click-through rate.
What Display Paths Do
Display paths appear after your domain in the ad URL:
www.yoursite.com/path-1/path-2
They do NOT need to match actual URLs on your site. They are purely cosmetic and help users understand what they will find when they click.
Display Path Strategies for Service Businesses
| Strategy | Path 1 | Path 2 | Best For |
|---|---|---|---|
| Service + Location | /plumbing | /houston | Local service businesses |
| Service + CTA | /emergency | /free-quote | Urgency-driven services |
| Service + Differentiator | /roofing | /licensed-pros | Trust-building |
| Service + Offer | /dental | /new-patient-99 | Promotion-driven |
| Service + Speed | /hvac-repair | /same-day | Emergency services |
Testing Display Paths
Display paths can be tested using the Ad Variations feature:
- Create a variation targeting the display path
- Test
/free-quotevs./get-estimatevs./book-now - Run for 2-4 weeks
- Measure CTR impact (display paths primarily affect click-through rate)
Small wins here compound. A 5% CTR improvement from display path optimization means 5% more traffic at the same budget.
The RSA + Quality Score Connection
While Ad Strength does not impact Quality Score, your RSAs still influence QS through the Expected CTR and Ad Relevance components.
How RSAs Affect Each QS Component
| QS Component | How RSAs Influence It |
|---|---|
| Expected CTR | Google predicts CTR based on your RSA's historical performance. Higher-CTR headline combinations improve this score. |
| Ad Relevance | Google evaluates whether your headlines/descriptions match the searcher's intent. Keyword headlines drive this. |
| Landing Page Experience | RSAs do not directly affect this, but message match between RSA and landing page matters. |
The Feedback Loop
Better RSAs > Higher CTR > Better Expected CTR score > Higher Quality Score > Lower CPC > More budget for the same spend > More data for testing > Even better RSAs
This is why the weekly testing cadence matters. Each improvement feeds the next.
QS Benchmarks for RSA-Driven Campaigns
| Quality Score | Status | Action |
|---|---|---|
| 8-10 | Excellent | Maintain, test incrementally |
| 7 | Good | Focus on landing page improvements |
| 5-6 | Average | Review headline relevance, test new combinations |
| 3-4 | Below average | Audit keyword-headline match, check landing page |
| 1-2 | Critical | Likely mismatch between keywords, ads, and landing page |
The average Quality Score across all advertisers is 6.2 out of 10. A one-point improvement reduces CPC by approximately 16%.
What Comes Next
RSA optimization is one piece of the conversion puzzle. Your ad copy is only as good as the page it sends traffic to.
Next in this series: Landing Pages for Service Businesses: Message Match, Single CTA, and the 1-Second Rule covers how to build landing pages that convert the traffic your RSAs generate.
Key Takeaways
The RSA Optimization Framework in 5 Points
- 2 RSAs per ad group — The data-proven conversion sweet spot from 93,055 ads studied
- 15 headlines, 4 descriptions — Fill every slot, but put keywords in only 20-30% of headlines
- Ad Strength is irrelevant — It does not impact Ad Rank, Quality Score, or actual performance
- Pinning is a trade-off — 3.9x fewer impressions but higher CTR and conversion rates. Use strategically.
- Test weekly using Ad Variations — One variable at a time, minimum 2 weeks, document everything
Priority Actions
| Action | Impact | Effort |
|---|---|---|
| Reduce to 2 RSAs per ad group | High | Low |
| Fill all 15 headlines + 4 descriptions | High | Medium |
| Stop optimizing for Ad Strength | High | Zero |
| Start weekly Ad Variations testing | Very High | Medium |
| Implement headline category framework | High | Medium |
| Set up testing documentation | Medium | Low |
What Doesn't Matter
- Ad Strength rating (compliance metric, not performance metric)
- Google's recommendation to add more keyword headlines
- Having 3+ RSAs per ad group (dilutes data)
- "Optimize" ad rotation setting (prevents proper testing)
What Actually Matters
- Conversion rate and cost per conversion (your money metrics)
- Headline diversity across 5 categories (keyword, benefit, CTA, trust, differentiator)
- Controlled testing with isolated variables (the only way to learn)
- Weekly review cadence (compounding improvements over time)
This guide is part of the Google Ads Efficiency Playbook 2026 series. Data sourced from Optmyzr (93,055 RSA study across 13,671 accounts), Lunio, PPC Mastery, Search Engine Land, and WolfPack Advising.