RSA Copy and Search Assets, Written Spanish-First

A Google responsive search ad holds fifteen headlines of thirty characters and four descriptions of ninety. In Spanish that is a harder brief than in English: the same sentence runs roughly a fifth longer, so a line sitting at twenty-eight characters in English arrives at thirty-four in Spanish and never serves. We write the Spanish set first and fit the English to it. Accounts built the other way round are why so much Spanish search copy reads like a customs declaration. What arrives: For each ad group in scope, the full asset set — headlines, descriptions, sitelinks written to twenty-five characters with both thirty-five character description lines, callouts, structured snippets — and beside it a sheet naming the claim each headline carries, where that claim is sourced, and the pin position it holds if any. Nothing goes in because a slot was empty. Where the copy comes from: The search terms report, not the keyword plan. We read the queries that converted over the last ninety days and the queries that spent without converting, and we ask for recordings or transcripts of around twenty sales calls. In industrial accounts the converting query is usually a part number, a standard, a material or a tolerance, and the headline that wins is the one repeating it back with the figure attached. In kids and family and in sports the converting query is a size, an age band, a fit or a return window. Copy that answers the query is not a style choice, it is the entire job. Pinning and Ad Strength: We pin. Ad Strength is not part of Ad Rank, and an Excellent rating bought by unpinning a compliance line or a price qualifier is a bad trade. If your internal reporting requires the rating, say so at the brief stage and we will tell you exactly what it costs you in control. On measurement, honestly: Asset reporting gives you Low, Good and Best labels with impressions, and no conversions by asset. Headline-level return cannot be read off the interface, whatever anyone tells you. When a claim matters enough to be worth knowing, we isolate it in its own ad group or an experiment, so the claim goes on trial rather than the asset. We will tell you which two or three claims deserve that treatment and which do not. Spanish specifics: Tú or usted is decided once, with you, and held across the account. Castilian for Spain. If you also sell into Mexico or Colombia we write those separately or not at all, never with neutral wording that satisfies neither. Legal and price qualifiers go into unpinned positions in both languages so they survive whatever the system assembles. Not included: Bidding, budget allocation, keyword expansion, landing pages, Performance Max asset groups, display. We will point at a landing page that contradicts an ad, once. If it stays, we stop writing that ad rather than write around it. Not for you if: You want dynamic keyword insertion running across the account — we strip it outside narrow branded cases and will not build on it. You measure copy work by assets produced per month. You need an agency of record. Or the account records under a few hundred conversions a quarter, in which case copy is not your constraint, and we will say so before invoicing.