See POAS vs revenue-only reporting
Profit Bid connects store costs to ad spend so you bid on margin — not vanity ROAS.
Credit for one sale (linear model)
- Display (view)20%
- Paid social25%
- Shopping click30%
- Brand search25%
A linear model spreads credit evenly. Switch models and the 'winning' channel changes overnight.
Track POAS automatically from your store — upload profit conversions and scale winners with A/C/X labels.
Common attribution models
Each model answers 'who gets credit?' differently:
- Last-click: 100% to the final touch — simple, biased to bottom funnel.
- First-click: 100% to the first touch — biased to top funnel.
- Linear: equal credit to every touch.
- Data-driven: credit based on modeled contribution.
Attribution's profit blind spot
Every model splits revenue credit, not profit. A touchpoint credited with a $200 revenue order might have driven a thin-margin, heavily discounted sale.
Combine attribution with contribution margin so you evaluate channels on the profit they influence, not just the revenue they are assigned.
Frequently asked questions
Common questions about this topic — tap to read answers.
Which attribution model is best?
Data-driven is usually most balanced, but any model is a simplification. Validate with incrementality tests before trusting it for budget.
How does iOS/privacy change attribution?
Signal loss makes deterministic attribution harder, pushing platforms toward modeling. This raises the value of first-party profit data.
