Shapley Value Attribution (Concept Demo)
See how Lloyd Shapley's 1953 game-theory formula splits credit across channels by marginal contribution — worked on a toy 3-channel example.
live — runs in your browserOnlinePost-campaign
Inputs
Results
This is the same Shapley value used in cooperative game theory since 1953 — public-domain mathematics, taught in every game theory course. Production ad platforms (see the patent reference below) apply it to millions of live touchpoint sequences in real time with additional machine-learned value functions; this demo applies the textbook formula to seven numbers you type in, so you can see the mechanism itself: each channel's credit is its average marginal contribution across every possible order channels could have been added to the path. That's why it differs from linear (equal-split) and last-touch, both of which ignore order and context entirely.
How this is calculated
Channel A — Shapley value
Source: Shapley (1953), Contributions to the Theory of Games
Channel B — Shapley value
Channel C — Shapley value
For comparison — equal/linear split (all channels)
What a naive equal-credit model would assign to each channel — same number for A, B and C
For comparison — last-touch model (channel C)
What last-touch attribution would give the final channel in the A+B+C path — all of it, zero to A and B
Every model runs locally in your browser. Nothing you type is sent anywhere.
Frequently asked questions
Is the Shapley Value Attribution (Concept Demo) free?
Yes. It runs instantly in your browser with no signup, and the methodology behind it is documented on this page.