Attribution

Credit for every touchpoint, labeled for what it is.

Seven credit models over one stream of touchpoints your organization owns. See how much of a channel's credit depends on the model before anyone argues about it.

One journey · 4 touchpoints · 1 conversion

Paid searchEmailOrganicPaid social

Seven models read the same journey and hand out credit differently. Pick per report; the underlying touchpoints never change.

Credit by model · same journey

Last Touch0 · 0 · 0 · 100
First Touch100 · 0 · 0 · 0
Linear25 · 25 · 25 · 25
Time Decay10 · 15 · 25 · 50
U-Shaped40 · 10 · 10 · 40
W-Shaped30 · 30 · 10 · 30
Learned MTA38 · 17 · 20 · 25

Seven credit models

Pick the model per report.

Every model reads the same touchpoints, so switching models changes the split, never the data.

  1. Last touch

    All credit to the final touchpoint. The number ad platforms report, so it's the easiest to reconcile against them.

  2. First touch

    All credit to the first touchpoint. Shows which channels start journeys, before the rest of the path gets a say.

  3. Linear

    Equal credit to every step. The baseline the shaped models are read against.

  4. Time decay

    More credit the closer a touch sits to the conversion, on a half-life you set. Suits promotions and seasonal pushes.

  5. U-shaped

    Forty percent to the first touch, forty to the last. For lead generation, where discovery and the close matter most.

  6. W-shaped

    Thirty percent each to the first touch, the lead milestone and the conversion. For B2B pipelines with a defined lead stage.

  7. Learned multi-touch

    Credit learned from your own converting and non-converting journeys. With too few conversions it falls back toward linear, and says so.

Built on data you own

Your journeys, your names, your window.

Identity and consent first

Identifiers resolve to people your organization owns. Consent, suppression and erasure apply before a journey is assembled.

Your own vocabulary

Channels, units and segments are yours to name. Renaming a channel relabels the report without rewriting history.

A lookback that fits

Set the window from one day to two years. It starts at 90 days. Longer journeys are counted as long, never cut off quietly.

Runs you can reopen

Every run is stored by model, so a figure quoted in a deck can be traced back to the run behind it.

Attribution and the model

Attribution describes. The model measures.

Credit models show which touchpoints sat on the path to a sale. What a channel actually caused, and what its next dollar will return, comes from the media mix model. Bayesin shows both side by side, and keeps them labeled.

See where the credit goes, and where the revenue came from.

Early access is by application. We set up each team ourselves.