Attribution
Trailing 52 weeksSwapping last-touch for a data-driven model moves Paid Search's credit by -9.2pt. The channels whose value moves most between models are the ones that will be argued about.
Every model below sees the same journeys and disagrees about who gets the credit. The question is not which is correct, but how much your answer depends on the choice.
Rule-basedModeled
Credit from modeled incremental lift per touchpoint.
Attributed revenue — data-driven
Trailing 52 weeks.
Modeled
Model sensitivity
How each channel's share shifts against a last-touch baseline. Large gaps mean the channel's value is model-dependent.
| Channel | Revenue | Share | vs last touch |
|---|---|---|---|
| Paid Search | $45.9M | 44.1% | -9.2pt |
| Paid Social | $21.7M | 20.9% | +3.4pt |
| Display | $7.9M | 7.6% | +3.0pt |
| $7.4M | 7.1% | -0.3pt | |
| Affiliate | $6.8M | 6.5% | +0.4pt |
| Video | $6.5M | 6.2% | +3.1pt |
| Organic Search | $5.6M | 5.4% | +0.3pt |
| Direct | $2.3M | 2.2% | -0.7pt |