Measurement first, then four channels in 90 days

The account was buying traffic, but nobody could prove what that traffic was worth. Conversion events were misfiring, double counting, or missing. Meta was optimizing on a pixel-only signal with no Conversions API backup. Amazon Attribution was not configured. There was no Microsoft presence at all. Every bidding algorithm is only as good as the conversion signal you feed it. Scaling budget on top of broken measurement multiplies the error, not the return.

Lightspeed eCom C-Series was a real constraint on structured data control and checkout flexibility. Migration was deliberately left out of scope rather than bundled in. A platform move is a separate decision, made because the current platform is capping growth, not because an agency prefers a different stack.

Week one went to tracking, not campaigns. All four paid channels ran inside the term, but not from day one. Search captured demand that already existed, so it moved first. Meta and Amazon needed the learning phase. Organic work started in week one so it could mature inside the same 90 days. The Amazon fulfillment question was audited before it was built.

Google Tag Manager and GA4 were validated event by event. The Meta pixel was corrected and the Conversions API connected for server-side redundancy. The Bing pixel and Microsoft Clarity were installed. Amazon Attribution was configured where the catalog supported it. Revenue in GA4 had to reconcile with Lightspeed closely enough to defend, not apologize for.

Google Ads Search and Shopping were restructured and scaled first. Microsoft Ads Search and Shopping were a new build for cheap incremental reach. Amazon Ads ran Sponsored Products and Brands. Meta ran Facebook and Instagram prospecting and retargeting, slowest to mature and best for new demand. Sequencing by intent lets the fastest-returning channels fund the patience the slower ones require.

Priority product and collection pages were optimized against high-intent keywords. Meta titles and descriptions were rewritten across core store pages and priority products. A technical sweep on C-Series addressed site structure, indexing, and crawl health. FAQ and answer-format content was added so the store could surface inside AI answer engines. Google Business Profile was refined for the Fredericton storefront. Paid stops the day the budget stops. Organic and AEO gains persist, which is why they belong in the same 90 days rather than a later phase.

The included discovery audit covered catalog and channel readiness, analytics and tracking health, and Amazon fulfillment automation feasibility across Multi-Channel Fulfillment, order routing, and FBA logic. Automation was found viable for a share of direct-to-consumer volume, with the rest staying on in-house shipping because of size, margin, or SKU eligibility. The build was not started. It was scoped and quoted separately once feasibility was proven, which is what stops a client paying for an automation their catalog margins cannot support.

Once paid and SEO were stable, remaining retainer hours went to international expansion and logistics groundwork: cross-border shipping cost modelling, duty and tax handling, and a first read on which US regions the catalog could serve profitably. Unused hours are budget the client already paid for.

Tracking was validated before spend increased. Google, Microsoft, Amazon, and Meta ran in sequence rather than all at once. SEO started in week one. Amazon fulfillment automation was answered as a feasibility question, not a surprise build. Return on ad spend is published here because the ad platforms reported it directly, measured over 30 days rather than picked from a good week. Revenue, ACOS, and ranking figures stay unpublished until GA4 and Search Console replace the internal draft. This model fits a catalog with depth, a functioning storefront, and enough margin to support paid media. It is the wrong starting point when the store needs replatforming first, when product content is too thin to convert bought traffic, or when unit margin cannot absorb ad costs. Media budget is paid directly to each platform by the client. C-Series remains the constraint it was on day one.