Cross-Channel Marketing Audit

Reactivated Feb 3, 2025 (after 13 months dormant) at a conservative $22/day, Maximize Conversions. Northline Digital scaled the budget hard over six weeks: $22 → $35 (Feb 20) → $65 (Mar 5) → $120 (Mar 22) — roughly 5.5× in a month, while ROAS held strong (9.40× in March).

Mar 25: switched to Maximize Conversion Value with an extremely conservative 900% Target ROAS — telling Google to only spend where it expects a 9× return. April’s ROAS came in even higher (14.47×), but on lower spend ($780 vs March’s $1,120) — a sign the target was capping volume even as efficiency stayed excellent.

That likely explains what came next: May 8, budget cut 60% ($120 → $48) — probably a correction once it was clear the tight ROAS target meant the bigger budget wasn’t being spent, not a vote of no confidence. May 15 eased the target to 600%, budget nudged up. May 28 (days before this audit) — a bigger move: Target ROAS cut to 350%, clearing room to spend more. June’s result — 7.17× ROAS on $1,510 spend — is that loosening showing up: lower ROAS than April, but the campaign is finally spending closer to its allocated budget.

An original PMax campaign was created the same day as Brand Search (Feb 3, 2025) at $18 budget. It appears only once in the change log — no further logged activity, and no performance data anywhere in the report under that name. It’s worth getting the specific policy-violation reason directly from the client and agency, since a repeat violation would hit whatever inherits its targeting.

A replacement was built from scratch on May 20 (new campaign, fresh history) — budgeted $40, cut same-day to $18, set to the same conservative 600% Target ROAS pattern as Brand Search. May 24 — after 4 days of data — budget raised more than 2.5× to $52, suggesting early results justified more spend. May 28 brought the same Target ROAS cut to 350% as Brand Search, in the same session, by the same user — a coordinated call across both Shopify campaigns, not an isolated PMax tweak. June finished at 6.85× ROAS on $410 spend — good, but it’s early days of a rebuilt campaign, not a trend yet.

The account’s largest single campaign by spend. Held a healthy 3–5× ROAS range through most of the window, then dropped to 0.95× in February 2025 — directly following a 90-keyword purge across an 8-day window (Feb 8–14) that hit this campaign hardest. Recovered to 4.1× by April as keywords were gradually re-enabled, but has faded again through May–June with no major structural change logged — most likely normal seasonal softening rather than a new execution problem.

Launched steadily but went completely dead for six weeks in March–April — zero conversions despite $60–90/week in continued spend. No status or keyword change in the log explains the dead spell; worth confirming this was a tracking or attribution gap rather than a real demand problem before writing the SKU off.

The most recently launched Amazon campaign and the weakest so far. Opened below breakeven, improved modestly through May, still hasn’t cleared 2× ROAS in any full month. Smallest and least proven of the three Amazon campaigns.