Which Channel Was Actually Working? Solving a 3-Platform Attribution Disaster

The situation

A Canadian beauty brand was running on Meta, Google Ads, and TikTok simultaneously. Their combined ad spend was $25K/month. Each platform's dashboard showed strong ROAS — but when the founder added up all the "attributed revenue" across the three platforms, the total was 3.1× their actual Shopify revenue. Every platform was claiming credit for the same sales.

The problem

Without a unified tracking system and consistent attribution rules, every platform was operating in its own bubble. A customer could click a Meta ad, see a Google remarketing ad the next day, and convert — and both Meta and Google would claim 100% credit. TikTok was doing the same. There was no way to know which channel was genuinely driving revenue, which meant budget decisions were being made entirely on self-reported platform data — the most biased data possible.

What we did

1. Built a unified GTM Server-Side container — all three ad platforms now route events through a single server endpoint with consistent data
2. Standardised UTM parameters across every campaign on all three platforms — consistent naming convention for source, medium, campaign, and content
3. Implemented cross-platform deduplication logic so each purchase is attributed once, following a consistent last-touch model
4. Built a Looker Studio dashboard pulling from GA4 as the single source of truth — showing true channel contribution independent of what each platform self-reports

Why it worked

Every ad platform is incentivized to claim as much credit as possible — that's how they justify your spend. The only way to get honest attribution is to measure from a neutral source (GA4 + UTMs) and enforce consistent cross-channel deduplication. Once the true channel split was visible, the answer was obvious: reallocate heavily toward Meta, scale back TikTok significantly. ROAS improved within the first month.