Search Account Audit and Restructure, Retail and Marketplaces

The deliverable is a restructure you can act on in a week, not a scorecard. We take one Google Ads account and its Microsoft Ads mirror, spend three weeks inside them, and hand back a rebuild with the campaigns, budgets, bid strategies and feed changes already specified. We take accounts spending 80,000 USD a month or more, because below that the six-week learning period a rebuild costs is worth more than the structure it buys. Where we start: Not in the account. In your order table. We join Google's click identifiers, gclid, gbraid and wbraid, to your orders, refunds and cost of goods, and rebuild the account's reported performance on contribution margin after landed shipping and ninety-day returns. In retail and marketplace accounts this usually moves the answer before we have touched anything: two campaigns sitting at the same reported ROAS separate by twenty points of margin once returns land, and the restructure follows that split rather than the ROAS column. What we actually check: Whether the conversion action driving Smart Bidding carries order value or a proxy that flatters cheap orders. Whether offline conversion import is running and how stale the last upload is. Whether Performance Max has brand exclusions set, and what its search term surface really shows once you look. Whether shopping listing groups are subdivided by anything that matches how you make money, or by product type because that is the default. Whether shared budgets are throttling the campaigns you want and feeding the ones you do not. Whether item availability, price and identifier coverage in Merchant Center match what your storefront is serving right now, because most Shopping waste we find is a feed problem wearing a bidding costume. On the Microsoft side, whether the import is a live copy or has been left to drift for a year. What we do not do: We do not audit an account behind the incumbent agency's back. If someone else runs this, they are told we are in it before we start. Procurement can use the findings afterwards; we will not be the anonymous knife. We do not restructure an account nobody will then leave alone, because a rebuild needs a four to six week freeze to exit learning, and a team editing daily through that window will read the rebuild as a failure and will have paid for noise. We do not rewrite ad copy here; this is a structure engagement. And we do not deliver a keyword expansion list, which is the cheapest thing in this trade to produce and the least likely to be your problem. Who this is not for: Lead generation accounts whose CRM carries no closed-won value. Accounts inside a bid management platform whose owner will not give us the strategy layer. And anyone who wants an audit to settle an internal argument about attribution models. We reconcile to the revenue your finance system reports, and if the two numbers disagree by more than five percent we stop and say so rather than optimise against a figure nobody trusts. How the rebuild lands: Staged. Structure and conversion plumbing first, then budgets, then bidding, a week apart, so when something moves you know which change moved it. The change log is written as we go and it is yours. We stay on for a fortnight past launch to read the learning period with you, and we name the decisions of our own that are not working before you find them. What we will not promise: A percentage lift. The account is one input, and the two that outrank it, your price against the market and what happens after the click, are yours.