Google and Microsoft Ads Audit Against Booked Revenue

Two thirds of the accounts we are handed are bidding toward a conversion the finance team would not recognise: a form submit, a call over thirty seconds, a GA4 key event on a thank-you page. Smart Bidding does exactly what it is told, so the account is optimised, correctly, toward the wrong thing. This audit re-scores every campaign, ad group and search term against what the business actually books, and tells you what that re-scoring is worth per month at your current spend. What we read first: The conversion set, before the campaigns. We list every active conversion action, its counting rule, its attribution window, whether it is primary or secondary, and whether enhanced conversions or an offline import feeds it. We reconcile the platform's conversion count against your CRM for the same ninety-day window and show you the gap in both directions. In Spanish and other EEA accounts we separate observed from modelled conversions under Consent Mode v2, because a bidding target set on a modelled base moves when your consent rate moves and nobody is told. Where the money usually is: Location settings left on Presence or interest rather than Presence, quietly buying a foreign audience. Lost impression share split into budget and rank, because those two findings have opposite fixes and get conflated in most reports. N-gram analysis across twelve months of search terms, which surfaces waste a term-by-term review never will, since the loser is rarely one query but a token appearing across four hundred of them. Performance Max broken out by channel through the account's own placement, asset group and channel reporting, so you can see how much of that return is search traffic wearing a costume. Portfolio bid strategies whose target has not been touched since the seasonality changed. Brand campaigns running without a holdout. What you receive: A written diagnosis ordered by monthly euros at risk rather than by a severity label, because severity labels cannot be prioritised. Every finding carries the report, the query or the API call that produced it, so your team can reproduce it without us. A ninety-minute call with whoever will implement, recorded. A one-page decision list you can hand to an in-house buyer or to an incumbent agency without editing it. Not included: We do not implement. The audit is read-only: standard access on your MCC, viewer on GA4, and a CRM export. We do not write ad copy, rebuild landing pages, audit SEO, or model attribution across channels you did not ask about. We do not benchmark you against anonymised industry averages, which describe other people's offers rather than your account. Who this is not for: Accounts under roughly forty thousand euros a month in media. Below that the findings are real and small, and your money does more on the offer than on the account structure. Advertisers who cannot share revenue data, because without it this becomes a settings checklist and any reseller will give you one of those free. And anyone commissioning the audit as ammunition to remove an agency: we will describe what the account does and does not do, we will answer factual questions from both sides, and we decline requests to phrase findings for a procurement argument. We read and report in English or Spanish.