How to Write Responsive Search Ad Headlines From Your Search Terms

Mine the exact phrases real customers typed, promote the ones that convert, and feed them into your RSA so Quality Score and relevance both climb.

October 8, 2026 · QueryCut

Pull your converting search terms from Campaigns > Insights and reports > Search terms, sort by conversions, and turn the exact phrases customers typed into RSA headlines. Doing this raises ad relevance because the headline now literally matches the query, which improves expected click-through rate and Ad Rank. The skill is not copying terms blindly: it is picking the ones with enough data and real intent, then shaping them into headlines that still read like a sentence.

Why search-term headlines work: the relevance mechanism

Google scores each eligible ad on expected CTR, ad relevance, and landing page experience. Ad relevance measures how closely your ad matches the intent behind the query. When a headline contains the exact words a user typed, Google can bold that match in the live ad, and the user sees their own language reflected back. That lift in perceived relevance is the whole point.

Your search terms report is the only place where you see what people actually typed, as opposed to what you bid on. With broad and phrase match, the gap between your keywords and real queries is large. A keyword like "crm software" can trigger on "crm for small construction business," and that longer phrase is a better headline than your keyword because it carries the modifier the customer cares about.

There is a reporting wrinkle worth knowing: Google hides low-volume and privacy-sensitive queries, so the search terms report is a sample, not the full picture. Headlines you seed from it are built on the visible, higher-volume portion of demand. That is fine, because those are the queries worth optimizing for anyway, but do not assume the report shows everything.

The pull: where to look and what to grab

Go to Campaigns > Insights and reports > Search terms. Set the date range to the last 30 to 90 days so you have enough volume. Add the Conversions and Cost columns if they are not showing, then sort by Conversions descending. The terms at the top are your headline gold: these are queries that both attracted clicks and produced results.

Also scan the terms sorted by Clicks, not just conversions. High-click, zero-conversion terms split into two groups: genuinely irrelevant junk (which you negative out) and relevant queries where your ad or landing page is simply weak. The second group is a signal that the intent exists but your messaging missed it, which is exactly the case a new headline can fix. The same report that surfaces headline candidates also surfaces the junk you should exclude, so do both passes in one sitting.

Export the list or copy the top 20 to 40 terms into a sheet. You are looking for recurring nouns, modifiers, and phrasings: words like "free," "near me," "for small business," "24 hour," a specific brand, a city, or a use case. Those modifiers are the raw material.

Decision rules for which terms to use

Not every term deserves a headline. Use these heuristics so you are promoting signal, not noise.

A term with at least 1 conversion and roughly 100-plus clicks over your window is a strong candidate: the intent is proven and the volume is real. A term with conversions but under 30 clicks is promising but thin, so treat it as a maybe and watch it. A high-click, zero-conversion term that is clearly relevant to your offer is worth testing as a headline, because the problem may be message match rather than demand. The reason for the roughly 100-click floor is simple: below it, a single missed or mistracked conversion swings your read, and you can burn budget chasing a phrase that never actually paid off.

Cut anything that is off-intent regardless of clicks. If "free" dominates your top terms but you sell a premium paid product, a "free" headline will raise CTR and tank conversion rate, which is worse than no headline. Match the term to your actual offer, not just to its traffic.

• Use: 1+ conversion and ~100+ clicks in the window, intent matches your offer.
• Test cautiously: conversions present but under ~30 clicks.
• Test as a message fix: high clicks, zero conversions, but clearly relevant.
• Reject: off-intent modifiers ("free," "jobs," "diy") that misrepresent your offer.

Shaping a term into a headline that reads right

RSA headlines cap at 30 characters each. Many search terms are longer, so you will trim. Keep the modifier that carries intent and drop filler. "affordable crm for small construction companies" becomes "CRM for Construction Firms" (26 characters) or "Affordable Construction CRM" (27). The customer's distinguishing words survive; the sentence stays grammatical.

Fill all 15 headline slots and all 4 description slots. RSAs mix and match, and Google needs variety to test. Seed 3 to 5 headlines directly from your top converting terms, then round out the rest with your value props, offers, and a call to action. Do not make all 15 variations of the same phrase, because that starves the test and can trigger a low Ad Strength rating.

Pin sparingly. Pinning a headline to position 1 guarantees it shows but removes it from rotation testing, so Google learns less. If you must guarantee your best search-term match appears, pin two or three alternatives to the same position rather than one, so Google still has choices. Reserve hard pins for compliance or brand requirements.

The failure mode: chasing the term instead of the intent

The trap is treating the search terms report as a copy-paste source. The most common mistake is promoting a high-traffic term that converts poorly because its intent is adjacent to yours. A plumber who sees "plumbing salary" with heavy clicks and writes a headline around "plumbing" generically will attract more of the wrong clicks. The term had volume; it did not have buying intent. Read what the person wanted, not just the words.

A second failure: over-fitting to one quarter's language. Seasonal and trend-driven phrasing fades. A headline built on "2024 tax deadline" is dead in February of the next year. Prefer durable modifiers (use case, audience, location) over time-stamped ones unless you plan to rotate them.

You also cannot see everything. Google withholds a slice of queries, and conversion data lags because of attribution windows. A term that looks like a zero-converter today may show conversions in a week. That is why the click floor and the date window matter: they are your defense against acting on an incomplete read. Tools like QueryCut automate this pull and flag both headline candidates and negatives in one pass, but the report itself in Google Ads gives you everything you need to do it by hand.

Questions people also ask