Clean takeover and audit. We assumed control of the Google Ads and Microsoft Ads accounts, mapped exactly where spend was going, and identified the generic, order-value-only Performance Max structure and the blended brand and acquisition demand as the core efficiency leaks. Themed Performance Max, built around the catalog. Instead of one broad high-value and one low-value campaign, we rebuilt Performance Max around the actual product lines that matter to buyers. This let budget, creative, and signals concentrate by product type, so each line could be read and optimized on its own merits rather than averaged into a generic pool. Brand separated and protected. We pulled Brand Search into its own lean campaign so we could defend the brand term cheaply against competitors bidding on it, and so acquisition performance was no longer flattered by brand conversions. Brand Search became the single most efficient line in the account. Demand Gen for the top of funnel. We added Demand Gen to reach cold, in-market dog owners at scale and feed the lower-funnel campaigns, a far better fit than the inherited display remarketing that had been doing little. Tightened acquisition and competitor layers. Dynamic Search and a disciplined competitor campaign rounded out coverage, kept efficient and capped so they supported rather than drained the account. Microsoft Ads rebuilt in parallel. We applied the same structural logic to Bing, treating it as a real channel rather than an afterthought. Return on ad spend there more than doubled on a lower-competition surface where acquisition costs are friendlier. Connected to the full ecosystem. Paid search does not work alone here. It feeds a BigCommerce store wired into a mature Klaviyo email and SMS program, a Smile loyalty engine, and a branded app. New paid customers enter a system designed to bring them back, which is what makes aggressive acquisition economically rational.