The Challenge

Revenue-based optimization was pushing spend into high‑gross but low‑margin products and markets. Costs lived outside Shopify (handling, packaging, marketplace fees), split across manual sheets per store. Teams couldn’t see which campaigns truly paid back after all costs, and GA4 had no signal beyond revenue.

What we built

• Ingestion from four stores + costs

• Pulled orders, refunds, items, discounts from each Shopify store.

• Ingested handling costs and various fees from structured spreadsheets; standardized formats and currencies.

• Multi-layer warehouse modeling

• Raw → staging → intermediate → profit marts.

• Reconciled orders with refunds/partial refunds; assigned per‑item COGS, shipping, payment, and handling fees.

• Calculated contribution margin and POAS at order, product, campaign, market, and day levels.

• Clean metrics layer

• Standard definitions for margin, POAS, MER, and payback.

• Tests for completeness, currency conversions, and accepted values to keep numbers trustworthy.

• GA4 + ad stack activation

• Sent profit and margin events back into GA4 with the right parameters.

• Enabled downstream bid strategies to optimize on profit proxies instead of revenue only.

Why this works

• One place computes true profit with all the messy costs included.

• Ad platforms receive a better signal, so budget shifts toward products and campaigns that actually make money.

• Finance and growth use the same definitions - no more revenue‑ROAS mirage.

What changed

• Budget reviews moved from “ROAS looks fine” to “POAS by campaign and product.”

• Low‑margin winners got capped; high‑margin sleeper products received more spend.

• Promo planning now considers contribution margin by market, not just top‑line.

Results

• Better decision-making on ads with POAS as the north star.

• Spend reallocated toward higher‑margin products and regions.

• Cleaner handoff between growth and finance; fewer end‑of‑month surprises.