Blog / August 10, 2026

Writing Google Ads copy that gets chosen, not just clicked

Any competent copywriter can raise your click-through rate. Write something vague and flattering, promise a low price you do not actually offer, and people will click. Then they land, read the real offer, and leave. You paid for all of it. The job of a search ad is not to be attractive to everyone. It is to be obviously right for one kind of person and obviously wrong for everyone else.

The reader is comparing, not reading

Nobody reads your ad in isolation. On a commercial search there are three or four paid results, an AI answer above or around them, and the organic results below. Your ad is read in a stack, in about a second and a half, against competitors who are all saying the same three things: quality, fast, trusted.

That is the whole game. Not persuasion, comparison. If every ad on the page says the same thing, the user picks the one at the top or the one with the lowest price. If yours is the only one that names a specific fact, it gets picked by the person that fact matters to.

Concrete beats superlative every time. Compare:

  • Professional roofing services, free quote, trusted local team.
  • Roof replacement in 3 days. Licensed, insured, 10 year workmanship warranty.

The second one loses clicks. It also loses the wrong clicks. Someone who wants a small repair sees roof replacement and moves on, and you keep that money.

Write to disqualify

The single most useful habit I have picked up from running large volumes of search accounts is writing at least one asset that pushes people away. A price floor, a minimum order, a service area, a business type. Something like: from $2,400. Or: for teams of 10 or more. Or: Varna and 30km around.

Advertisers hate this because the click volume drops and the dashboards look worse. But cost per click is a cost, not a metric. If your ad tells people your price range before the click, the people who click already accepted it. Your sales calls get shorter and your close rate goes up, and neither of those things shows in the Google Ads interface.

This works especially well for anything where the price surprises people. Custom work, medical, legal, construction, B2B software with a real implementation cost. If you have ever ended a sales call by saying the budget is much lower than what we do, your ads should have said the number.

How to build a responsive search ad that is actually testable

Responsive search ads mix your headlines and descriptions automatically, and asset level reporting now shows real click and conversion data per asset rather than a vague label. That only helps if you gave the system distinct things to compare. Fifteen headlines that all say the same thing produce fifteen rows of noise.

The structure I use:

  1. Three to four headlines that name the exact service and the exact place or segment. These carry the relevance. Pin one of them to position one if the campaign is tightly themed and drifting.
  2. Three headlines that are proof. Years in business, number of installs, warranty length, certification, response time. Facts you can defend if a customer asks.
  3. Two or three that are the disqualifier. Price floor, minimum, geography, or who it is not for.
  4. Two that are the action. What happens after the click. Not learn more. Book a 15 minute call. Get a fixed quote in 24 hours. See sizes and stock.
  5. Descriptions that answer the objection the reader has at that moment, which is almost always: will this waste my time, and what will it cost.

Then leave it alone for long enough to gather data. Judging asset performance on a handful of conversions is guessing with extra steps.

Assets do most of the persuading now

The text ad is a small part of the surface area. Sitelinks, callouts, structured snippets, images, prices, and call assets often take up more of the ad than the headlines. Call ads as a standalone format are being retired, so if your business runs on phone calls, the call asset attached to a normal search ad is where that now lives. Get it in place rather than waiting for the deadline.

Use sitelinks as navigation for different buyer states, not as a list of your menu. Pricing, service area, how it works, and a real case are worth more than about us and contact.

What I check before approving copy

  • Would a competitor be able to put their logo on this ad without changing a word? If yes, rewrite it.
  • Does the landing page repeat the exact phrase from the headline? If not, the click is already half wasted.
  • Is there one number in the ad? Price, time, quantity, years, distance.
  • Does it promise anything sales cannot deliver? Free is the worst offender. If the first consultation is free but the survey is not, say which.
  • Does it read like a person wrote it? Automated ad copy generation now pulls from your site, and it is competent but generic. Your job is to be the specific voice it cannot invent.

What to do next

Pull your top five ad groups by spend. Print the ads next to your competitors’ ads for the same query. If you cannot tell in one second which one is yours, that is your first rewrite. Add one disqualifying asset per ad group this week, and check cost per lead and close rate together in 30 days, not click-through rate alone. If you want a second pair of eyes on the copy and what it is costing you, you can book a free consultation.

← All postsBook a free consultation