Responsive search ad copy written to your conversion data

Ad copy is written here after the conversion definitions have been read, never before. In an ad group producing fewer than thirty conversions a month, no headline test resolves inside a quarter because variance swamps the effect, so this engagement is not test design. It is deciding what the ad is allowed to claim, writing it, and standing behind the claim when your legal or programme team asks who wrote that line. What we write: Fifteen headlines and four descriptions per responsive search ad, built as a set with combination logic in mind rather than as a list of interchangeable lines. Pinning is used sparingly and every pin carries a stated reason: a regulated claim that must appear, a brand name that must lead, a price that must not land beside an unrelated offer. Most accounts we inherit are pinned to the point where a fifteen-headline asset set is one ad in fifteen costumes. Unpicking that, and showing you the queries it was hiding, is the first day of work. Hindi and Hinglish: For Indian search we build Latin-script transliteration and Devanagari as separate asset sets. They attract different queries from different people, and mixing scripts inside one ad produces copy that reads as machine output in both. Which script should carry the account is a question about your buyer, and the answer is usually already sitting in twelve months of search terms. We read that before writing a line. Sector rules we hold to: Kids and family: no claim about safety, development or outcome that a parent could reasonably read as a promise, and no urgency mechanic aimed at parental anxiety. That rules out much of what performs, and we say so before you sign rather than after. Nonprofit: donation copy names what a specific amount buys, and the figure has to survive a look at last year's accounts. Fashion: discount-led headlines teach Smart Bidding to find discount buyers, and it keeps finding them after the sale ends, so where contribution margin matters more than volume we write around price and record the trade in the handover. What arrives: Asset sets for the ad groups that carry spend, not for every ad group in the account. A one-page rationale per campaign covering what each set argues and what it deliberately does not say. A rejected list, being the lines we drafted and cut with the reason attached, which is the part clients quote back to us most often. And a short note on which ad groups should be merged before the copy can be judged at all. Not included: Landing page copy, though we tell you plainly when the ad is writing a cheque the page does not honour. Performance Max asset groups, which are a different piece of work with different failure modes. Display, video and app promotion text. Dynamic keyword insertion and countdown customisers as a standing habit; both have narrow legitimate uses and we propose them by exception, never by template. Who should not buy this: Anyone who measures this work by the Ad Strength label. It reads asset diversity, not the argument the ad makes, and writing toward it produces vague copy that scores well and sells nothing. Anyone who wants a large number of variants per ad group to pick from, which is a shopping exercise and not this service. And any account without working conversion tracking, where nothing written here can be judged. That work comes first and it is a different engagement.