Attribution Model Calculator
The model decides the answer, not the data.
ROAS on last touch
0.68×
1.48× on first touch
The same channel and the same spend produce 1.48× or 0.68× depending purely on which model you pick. Attribution does not measure contribution — it allocates credit, and the allocation rule is a choice. Only a holdout test answers the actual question.
How the Attribution Model Calculator works
The same channel with the same spend produces wildly different returns depending purely on which attribution model you pick. Attribution does not measure contribution — it allocates credit, and the allocation rule is a choice rather than a finding.
Not financial advice. Marketplace fees change, and they vary by country, plan and seller status. Every rate here is an editable default, not a quoted price — check the platform's current fee schedule before you price a product against it. This is not tax or business advice.
Frequently asked questions
Which attribution model is correct?
None of them. Each encodes an assumption about how influence works, and no model can be validated from the same data it describes. They are useful for consistency, not for truth.
What is the difference between first and last touch?
First touch credits discovery; last touch credits the final click. Upper-funnel channels look excellent under first touch and terrible under last, and neither is measuring their real contribution.
How do I actually measure contribution?
Incrementality testing — turn a channel off in a matched region and measure the difference in total revenue. It is the only method that answers the question.
Should I still use an attribution model?
Yes, consistently, for optimisation within a channel. Just do not use it to decide budget allocation between channels without a holdout test to check it.