Attributiongraph Lite

Enterprise marketing teams can have sophisticated attribution and media-mix outputs but still depend on analysts to translate dashboards into answers about pipeline, spend, risk, and next actions. The supplied research confirms an attribution platform with a data export API, an enterprise analytical assistant category, and no existing product wrapping the named attribution outputs in a conversational interface with consent-aware refusal. Attributiongraph Lite reads only authorized model outputs, metric definitions, consent state, brand constraints, and source lineage. It classifies the question, identifies the governing model and period, calculates or retrieves the answer, cites the rows and transformations, exposes missing channels and uncertainty, and refuses when scope, identity, consent, freshness, or coverage is insufficient. A model estimate is not causal truth; a scenario is not a forecast; a suggested reallocation is not approval or execution; platform acknowledgment is not delivery, pipeline, revenue, or incremental lift. The product cannot autonomously move budget, create audiences, upload sensitive data, alter campaigns, or claim a future outcome. Every answer is versioned with model, population, assumptions, counterfactual, confidence or interval where available, and regression tests. Pricing evidence confirms an enterprise floor and usage-based analysis infrastructure, but the supplied buyer budget and deployment economics remain to validate.

The CMO, marketing analytics lead, marketing-finance partner, or data-product owner at a 1,000-plus-employee organization already using attribution or media-mix outputs.