How to fix: RSA headlines lack unique themes

TL;DR

Responsive Search Ads need headlines that cover distinct themes, not minor word swaps of each other. An RSA with fifteen headlines like "Buy Now", "Buy Today", "Buy Online", "Order Now", "Shop Today" feeds Google's combination engine one signal disguised as fifteen, collapses the auction-time variety the algorithm depends on, and gets flagged for repetition in Ad Strength scoring [1][2]. The fix is to map every slot to a unique theme — brand, primary benefit, secondary benefit, feature, social proof, offer or CTA, urgency, locality — and rewrite duplicates into themes you do not currently cover. Aim for fifteen distinct themes. Google reports a 15% lift in clicks and conversions on average when advertisers move RSAs from Poor to Excellent Ad Strength [1].

Why it matters

The RSA combination engine works by assembling headlines and descriptions into ad variants tuned to each user query. With fifteen headlines and four descriptions Google can theoretically build thousands of permutations, but only if those slots carry genuinely different messages. When five headlines are syntactic variations of the same CTA, the engine treats them as one bucket — the effective slot count drops from fifteen to whatever number of distinct themes you actually wrote. Ad Strength scoring explicitly penalises this: "Increase Headline and Description Uniqueness" is one of the specific signals Google surfaces when the system detects repetition [2].

The downstream effect is twofold. First, Ad Strength itself drops, and Ad Strength is one of the inputs to Quality Score's expected-CTR and ad-relevance components — see Ad Strength below Excellent for the broader mechanism. Second, Smart Bidding sees a narrower combination space and predicts lower conversion probability on the ad, which feeds into lower bids, lower impression share, and lower CTR in a self-reinforcing loop. Google reports a 15% average lift in clicks and conversions when advertisers move RSAs from Poor to Excellent [1]; theme uniqueness is one of the three factors scored.

Variety is not just for the algorithm — it is also how you find out what actually persuades your audience. If every headline says "Buy Now" you learn nothing from the asset report. If headlines span benefit, proof, urgency, and feature angles, the report tells you which angle drives clicks and which dies. That intelligence compounds across the account.

How to fix

1. Open the under-performing RSA. Go to the campaign, then the ad group, then Ads and assets → Ads. Click the RSA flagged with low Ad Strength or with the "make headlines more unique" hint in the right-hand panel.

2. Map current headlines to theme buckets. List the fifteen slots and tag each with one of these eight buckets:

   • Brand — the company or product name
   • Primary benefit — the main outcome the user gets ("Save 10 Hours a Week")
   • Secondary benefit — a different angle on value ("No Credit Card Required")
   • Feature or use case — a specific capability ("Built for Shopify Stores")
   • Social proof — numbers, awards, "trusted by" ("Used by 5,000 Teams")
   • Offer or CTA — the action you want ("Start Free 14-Day Trial")
   • Urgency — time or supply pressure ("Limited 2026 Pricing")
   • Locality — geo or language signal where relevant ("Serving London Since 2008")

   Two headlines per bucket is the ceiling, and the second must use a different angle, not a synonym.

3. Rewrite duplicates into uncovered themes. If you have "Buy Now", "Buy Today", "Order Now", "Shop Online" all sitting in the Offer/CTA bucket, four of those are duplicates. Pick the strongest one, then rewrite the other three as social proof, urgency, and feature headlines you did not previously have. The slot count stays at fifteen; the theme count rises.

4. Vary syntax as well as vocabulary. Two headlines can occupy different buckets but still read identical to the engine if both follow the same syntactic pattern. "Save 10 Hours a Week" (benefit) and "Save 30% on Costs" (offer) are syntactically twins. Break the pattern — make one a noun phrase, one a question, one a number-led statement.

5. Check descriptions for the same disease. RSAs allow four descriptions at 90 characters. If all four read like "Best-in-class service trusted by thousands. Contact us today.", you have the same theme problem at the description level. Apply the bucket exercise to descriptions too.

6. Save and verify. Ad Strength recomputes within minutes. The right-hand panel should move from Poor or Average to Good or Excellent. If it still flags repetitive headlines, the algorithm is telling you two slots are still too close — usually two CTAs or two benefits with synonymous verbs.

Common mistakes

• Counting words instead of themes. Fifteen filled slots with five buckets still scores as five themes. The system reads meaning, not character count.
• Synonym swaps as "uniqueness". "Buy Now" → "Purchase Today" → "Order Online" is one theme written three ways. The engine collapses these.
• Stuffing the keyword into every headline. This boosts asset-keyword relevance, the first Ad Strength factor, but tanks the diversity signal, the third. Net effect is usually a downgrade.
• Ignoring the right-hand hints panel. Google literally tells you which signal is failing ("Make your headlines more unique", "Add more headlines"). The hint is the diagnostic — read it and act on the specific failure.
• Refreshing only the worst RSA in the account. Repetition is usually a copywriting habit, not a one-ad accident. Audit all RSAs in the top-spend campaigns at the same time.

FAQ

Is eight unique themes enough, or do I need fifteen? Eight is the minimum Ad Strength likes to see for "Good"; fifteen pushes toward "Excellent" if the themes are genuinely distinct. The audit rule warns at fewer than eight unique headlines per RSA because below that line the combination engine has nothing useful to test.

Can I keep one duplicate if it's a strong CTA? One duplicate is fine — two CTAs covering slightly different intents (free trial vs demo, for example) read as one bucket but two angles. Three or more CTAs in the same bucket is the line.

Does "make headlines more unique" mean unique across the whole account? No. Ad Strength scores within the single RSA, so the comparison is across the fifteen slots of that ad. Different RSAs in the same ad group can repeat themes — though it usually makes sense to vary them anyway, because you are testing different ad angles.

What about Dynamic Keyword Insertion headlines? A DKI headline counts as one slot but its rendered text varies by query. Treat the underlying template as one theme (usually keyword-relevance) and write the other fourteen slots around it.

How often should I refresh? Audit Ad Strength quarterly. Refresh the lowest-CTR headlines from the Asset Report every 90 days; do not touch the high-performing ones. Repetition tends to creep back in over time as writers reach for safe CTAs.

Sources

1. Google Ads Help — About responsive search ads. RSA structure (15 headlines, 4 descriptions, 30/90 character limits), guidance to "provide as many unique headlines as you can", and the 15% clicks-and-conversions lift figure from Poor to Excellent Ad Strength.
2. Google Ads Help — About Ad Strength. Ad Strength scoring factors (asset-keyword relevance, headline/description quantity, sitelink sufficiency), the "Increase Headline and Description Uniqueness" recommendation signal, and explicit guidance to "avoid adding repetitive words, phrases, or ideas".