Shopify tells you what sold. It doesn't tell you why.

Shopify's analytics are designed for single-store operators who need basic sales reporting. The moment you need to answer questions like "What's the 90-day LTV of customers acquired through Meta vs. Google?" or "Which product categories have the highest first-to-second purchase conversion rate?" you've outgrown what Shopify can do. These questions require cohort analysis that joins order data, customer data, and acquisition source data over time windows that span months. Shopify doesn't store this in a queryable way. Your warehouse does.

The really painful part isn't that data lives in three platforms — it's that the same customer might exist in all three with no shared identifier. Jane buys on your Shopify site, then buys again on Amazon. Those are two different customers in your current reporting, but one actual human in your business. Your combined revenue, your true customer count, and your actual repeat purchase rate are all wrong because nobody has resolved the identity across platforms. A warehouse with proper customer stitching gives you the real numbers for the first time.

We've audited ecommerce ad accounts where the combined claimed ROAS across all platforms was 8x, but the actual blended ROAS — calculated from warehouse revenue divided by total ad spend — was 2.8x. The discrepancy comes from double-counting, view-through attribution, and each platform's incentive to over-report. The only way to get a real number is to pull all spend into the warehouse, pull all revenue into the warehouse, and match them with your own attribution logic. When you do, the numbers are lower — but they're real, and you can actually optimize against them.