How to Write High-Converting Google Search Ads: Ad Copy Playbook for DTC and Ecommerce Brands

This guide covers how RSA structure works, how to build a headline bank that holds up across match types and query intent, when pinning helps versus hurts, and how copy connects directly to CPC and ROAS.

How Google RSAs Work

RSAs replaced Expanded Text Ads as Google's standard Search ad format in 2022. You supply up to 15 headlines and 4 descriptions. Google tests combinations automatically, serving the arrangement most likely to get a click for that specific query and user context.

The combination pool is large: 15 headlines can theoretically produce 43,680 distinct three-headline arrangements. In practice, Google narrows this quickly using performance history. Under-performing combinations get served less. Strong performers run more.

The implication: every headline you write either adds to or dilutes your combination pool. Writing 15 slight variations of the same sentence does not give the algorithm much to learn from. Writing five distinct categories of headline does.

What Goes Into a High-Performing RSA

Think of your headline bank in terms of function, not just content. Each headline serves a different job in the ad.

Keyword echo: Mirrors the search query to show ad relevance. "Women's Trail Running Shoes," "Organic Protein Powder for Athletes." This type almost always belongs in the bank.

Benefit or outcome: Answers why someone should click. Not a list of features. "Feel full, not sluggish." "Runs that don't wreck your knees." Specificity matters here.

Differentiator: What separates you from the other ads on the page. Free shipping thresholds, subscription discounts, specific formulations. Vague differentiators like "premium quality" do not move CTR.

Social proof: Review count, certifications, press mentions. "18,000 five-star reviews" or "As seen in Runner's World." These compress trust into 30 characters.

Offer or urgency: Limited-time pricing, welcome discounts, bundle deals. Works harder during promotions but can run year-round when there is a real offer to show.

Descriptions do heavier lifting on conversion framing. Use them to expand on the offer, address a likely objection, and reinforce the landing page promise.

Building a Headline Bank for Ecommerce

Most accounts we inherit have 8-10 headlines that are variations of the same two ideas. The result is an RSA with high theoretical combinations but low actual diversity, and Google cannot extract a meaningful learning signal from 10 headlines that say essentially the same thing.

A functional headline bank for a DTC brand covers these seven types:

Write at least two headlines in each functional category. This gives Google real variation to test and avoids the situation where all 15 headlines are competing for the same job.

Pinning: When to Use It and When to Step Back

Pinning locks a headline to a specific position. Position 1 always shows; position 2 usually shows; position 3 shows less often. Pinning tells Google: this headline always appears here, regardless of what the algorithm learns.

The problem with over-pinning is reduced combination variety. If you pin three headlines to position 1, only one of the three rotates at a time. You have constrained what the algorithm can learn, Ad Strength drops, and Google's ability to optimize drops with it.

According to Google Ads Help, pinning reduces the overall number of headlines and descriptions that can be shown and limits the system's ability to test combinations, which is why Google does not recommend it for most advertisers (Google Ads Help, About Ad Strength for responsive search ads).

When pinning makes sense: legal disclosures, brand names with compliance requirements, regulated categories including supplements and financial products. Pin because you have to, not because it feels like more control.

Most ecommerce brands do not have a legal reason to pin. Let Google test which headline drives the best CTR for each query context. Restricting that with unnecessary pins costs you data and combination variety for no gain.

How Copy Connects to Quality Score (and CPC)

Quality Score is Google's per-keyword, per-ad rating. Three components: Expected CTR, Ad Relevance, and Landing Page Experience. Measured on a 1-10 scale, with 7+ considered strong.

It connects directly to cost. A high Quality Score means you pay less per click at the same position than a competitor with a low score. The exact CPC discount varies by auction, but the directional relationship is consistent: better copy equals more efficient spend at scale.

Expected CTR: Driven by headline pull. A keyword echo in position 1 generally boosts expected CTR because the ad visually confirms relevance to what the user typed.

Ad Relevance: Copy-to-query alignment. If someone searches "organic probiotics for dogs" and your ad serves generic pet nutrition headlines, relevance suffers. Tighter keyword-to-ad-group organization so copy can echo specific queries is the fix.

Landing Page Experience: Downstream from copy but connected. If your headline promises "free returns on all orders" but the landing page shows no returns policy, bounce rate rises and landing page score drops.

In our accounts, tightening headline-to-query relevance is one of the levers we use to lift Ad Relevance and Expected CTR, the two Quality Score components most driven by copy, without touching bids or budgets. Google Ads Help reports a related benchmark: advertisers who improve their Ad Strength from Poor to Excellent see 15% more conversions on average. Ad Strength is not the same as Quality Score, but both reward the relevant, varied copy this work produces (Google Ads Help).

Writing Copy That Converts, Not Just Clicks

CTR and CVR can pull in opposite directions. Broad, high-curiosity headlines drive clicks from buyers who may not be ready to purchase. Specific, offer-forward headlines drive clicks from buyers closer to the decision.

Include both types in the headline bank. Google will match high-curiosity headlines to broader awareness queries and offer-forward headlines to purchase-intent queries.

For queries with high commercial intent ("buy [product] free shipping"), headlines that confirm the transaction work best: specific offers, price anchors, shipping terms. For informational queries bleeding into your match types, benefit-led or educational headlines hold up better.

One area where DTC brands consistently underperform: dynamic keyword insertion. Inserting the query directly via {KeyWord: Default Text} can boost relevance scores, but it reads mechanically in ecommerce contexts where the query might be a long, awkward phrase. Test it against manually written keyword echoes before assuming it wins.

How We Test RSA Copy at jetfuel.agency

Google's Ad Variations tool lets you create a draft RSA with a changed headline bank, run both simultaneously, and compare CTR and conversion rate at statistical significance. This is more reliable than a before-and-after comparison, which can be affected by seasonality and bid changes.

The practical complication is volume. Google's machine learning needs traffic to settle. Low-volume ad groups can take 6-8 weeks to generate enough data to read a copy test cleanly.

For accounts with this constraint, we focus on the highest-volume ad groups first, run copy tests there, and apply learnings across similar groups rather than running parallel tests everywhere at once.

A few changes that consistently move the needle across ecommerce RSA accounts: replacing generic differentiators ("premium products") with specific ones ("certified organic, cold-processed"); adding real social proof counts when the number is meaningful, not just "thousands of customers" but "47,000 orders shipped"; removing headlines that duplicate the landing page headline exactly; and aligning description 1 to the landing page headline so the transition feels intentional.

For how copy strategy fits into the broader Google Ads structure, including PMax campaigns where creative assets work differently, see our Google Ads Performance Max strategy guide for ecommerce.

Frequently Asked Questions About Google Search Ad Copy