When we walk into a flat e-commerce account, the temptation — for everyone, including us — is to start moving bids, swapping audiences, and rebuilding campaign structure on day one. We have learned to resist it. Stagnant performance almost always traces back to one of three places: the creative, the post-click experience, or the data the platforms are seeing. Bid tactics and structure changes are downstream of those, and rearranging them first usually produces a week of motion followed by another flat month. The patterns below are common starting points across the retail and DTC accounts we have scaled, not guaranteed results — every brand has its own constraints — but they describe the order in which we test and what we have learned to test first.
The first 14 days are diagnostic, not tactical. We pull a 90-day window of platform data and back it against a ground-truth source — Shopify, the warehouse, or the order-management system — to confirm what is actually selling and at what margin. We audit feed health (titles, images, GTINs, availability, price parity across feeds), verify that the Conversions API or Enhanced Conversions is firing with deduplicated events, and walk every paid landing page on a real phone over a real cellular connection. More than half the time we find a fixable post-click issue — a slow LCP, a broken variant selector, a discount that does not apply at checkout, a shipping threshold that hides until the cart page — that is silently capping every channel. Those fixes go in before any media changes, because there is no point optimizing traffic to a leaky funnel.
Creative is where the real lift hides, and the lever is volume more than polish. In our experience, accounts plateau because they are running three or four ads on repeat for months at a time; the platforms have learned the audience and the audience has learned the creative. We typically launch a creative testing matrix structured around two axes: angle (problem-solution, social proof, founder story, comparison, demonstration, lifestyle) and format (static, short-form UGC video, motion graphic, carousel). The goal in month one is not to find the winner but to generate enough variants to give the algorithm room to work — usually 12–20 fresh assets per active campaign — and then double down on the angles that earn impressions and high thumb-stop rates. We pair the matrix with SKU-level winners from the audit so the spend lines up with the products that already convert and have the margin to support advertising.
Audience and structure changes come third, and they are usually narrower than expected. On Meta, broad targeting with strong creative tends to outperform tight interest stacks once the data is clean; we collapse fragmented ad sets into one or two well-fed ABO or ASC campaigns and use exclusions to manage overlap rather than building parallel structures. On Google, the unlock is most often Performance Max with a tightly merchandised feed and asset groups split by margin tier — alongside a non-brand search campaign that catches the demand PMax does not (and a brand campaign that is honest about its incrementality). We expand audiences in a deliberate order: existing customers and lookalikes first, broad next, retargeting layered with strict frequency caps, and connected-TV or YouTube only when the budget can sustain a meaningful weekly reach.
Measurement keeps the program honest. Platform-attributed ROAS is a useful diagnostic but a poor objective; we anchor reporting on blended marketing efficiency ratio (MER), incremental ROAS via geo or holdout testing, and contribution margin after platform fees and cost of goods. New-vs-returning customer split, cost per acquired customer, and the trailing 90-day repeat rate tell us whether growth is healthy or whether we are just buying repeat purchases at full price. By day 90 we expect a clear picture of which channel and which creative angle deserves more budget, which post-click fixes are still pending, and which audiences are saturated — and that picture is what drives the next quarter's roadmap. Want this run on your account? Call (888) 338-9816 or email onboarding@gofluxmarketing.com.