The Shopify Blind Spot: How 65% of Purchases Were Invisible for 18 Months

The situation

A US sports nutrition brand had been scaling Meta spend for 18 months. Their pixel reported 31 purchases per day on average. Their Shopify dashboard confirmed 89. They had assumed this was a normal attribution window discrepancy. When I explained what was actually happening, they were stunned.

The problem

Shopify's checkout process redirects users to a Shopify-hosted checkout domain (checkout.shopify.com or a custom domain Shopify controls). Browser-based pixels physically cannot fire on this page — it's a different domain entirely, and there's no way around this with a standard pixel implementation. Every customer who completed checkout on that hosted page was invisible to Meta. Over 18 months, Meta's algorithm had been building its entire lookalike audience and smart bidding model on a 35% sample of real customers.

What we did

1. Confirmed the root cause by mapping Shopify's checkout flow and identifying the exact point where pixel events stopped firing
2. Built server-side CAPI using Shopify's Order Created webhook — this fires from Shopify's backend for every completed order, regardless of which checkout page the customer used
3. Added full PII enrichment: email, first name, last name, ZIP code, and phone number for every event — maximizing the Event Match Quality score
4. Ran browser pixel and CAPI simultaneously with deduplication for 14 days, comparing against Shopify daily to confirm accuracy before switching to CAPI-primary

Why it worked

This is one of the most common — and most damaging — tracking gaps in Shopify stores. The browser pixel appears to be working in Event Manager because it fires ViewContent, AddToCart, and InitiateCheckout correctly. But the purchase event, which is the only one Meta's algorithm optimizes on, disappears the moment the customer reaches checkout. CAPI via Shopify webhook is the only reliable fix.