Attribution
Credit where the
touchpoints are.
Six credit models over a touchpoint stream your organization owns: Last Touch, MTA, Linear, Time Decay, U-Shaped, and W-Shaped, with identity resolution, lookback windows, and a lawful basis recorded per subject.
One journey · 4 touchpoints · 1 conversion
Six models read the same journey and hand out credit differently. Pick per report; the underlying touchpoints never change.
Credit by model · same journey
How it works
01
Touchpoints and conversions
Every interaction lands in one stream. Conversions join to it by resolved subject, not by a vendor's cookie.
02
Resolve identity
Identifiers resolve to subjects. Consent, suppression, and erasure are enforced before a journey is assembled.
03
Run credit models
Choose one of six models per report. Runs are versioned, so last quarter's credits stay reproducible after a rename.
What you get
Six models, one journey
Last Touch, MTA, Linear, Time Decay, U-Shaped, and W-Shaped read the same touchpoint stream, so you can see how much of a channel's credit depends on the model before anyone argues about it.
Org-defined channels, units, segments
Vocabulary is data. Renaming a channel relabels the report without refitting history.
Lookback that covers the real one
Set the window per organization. Journeys longer than the window are counted as such, not silently truncated.
Reproducible runs
Attribution runs and credits are stored per model, so a figure quoted in a deck can be reopened.
What it will not do
Boundaries we state up front, so the numbers are trusted.
- It does not claim causal impact. Credit models describe journeys; MMM and experiments estimate lift.
- It does not resolve identities you lack a lawful basis to process.
- It does not backfill history from sources that no longer have it.
Also in the product