Outcome tracking and holdouts
Outcome tracking and holdouts in iReveal: every decision carries its result in a single trace, and incrementality is measured against counterfactuals.
Analytics ends at conversions and e-commerce events; pipeline and revenue live somewhere else, and the connection between the two is asserted, not shown. iReveal closes that gap: every decision carries its result in a single trace, and the trace ends in pipeline and revenue impact, not pageview counts.
The trace
A trace is the full lineage from event to outcome: the visit that was captured, the identity it resolved to, the signal it produced, the decision that acted on it, and what came of that decision. Because the chain is one object, attribution is something you read, not something you model after the fact, and the record is audit-ready. The trace structure is the same one the full platform uses; see Decisions and traces.
Holdouts and causal measurement
Correlation is what last-touch attribution sells you: what coincided with the number moving. iReveal measures incrementality against counterfactuals: what actually moved the number. Holdouts are how the counterfactual exists at all, a portion of eligible records held back so the lift of acting can be read against the cost of not acting.
Alongside your MMM
Marketing-mix models and geo-lift tools measure channel-level lift on aggregate data, with no pixel and no user-level tracking, and they are strongest on channels like TV, CTV, and out-of-home. iReveal is trace-level: person-, SKU-, and decision-level causality that also recovers signal and activates.
Last updated