Intent Stage
A funnel-position descriptor for on-screen audiences — awareness, consideration, decision, post_purchase, unknown. Trillboards-emitted.
Intent stage is a Trillboards-emitted audience signal that classifies the inferred funnel position of the on-screen audience. The enum has five values: awareness, consideration, decision, post_purchase, unknown. The signal rides on Trillboards' segtax=600 namespace and is one of six declared cloud-emit fields shipped per playout.
The inference is multi-signal. The cloud Gemini call considers venue category (a sneaker store skews "decision"; a gym lobby skews "consideration"), time of day (lunch crowd at a quick-service restaurant skews "decision"), audience activity (commuting skews "awareness"; shopping skews "decision"), and what's visible (cart contents, brands worn). The output is the discrete bucket plus a confidence score; the raw inferences stay in the embedding layer.
For buyers, intent stage maps directly to creative selection. An awareness campaign wants screens with audience in the awareness bucket; a retargeting drip wants screens with audience in the decision or post_purchase bucket. Combining intent stage with cohort composition and attire archetype creates a buy-side targeting clause that's denser than anything traditional DOOH can offer.
Per Trillboards' emission contract, intent stage is conditional — emitted only when the cloud has enough signal to classify with confidence above the per-venue threshold. Below threshold, the field comes back as "unknown" and the buyer's targeting logic should fall back to venue-category targeting.
Authoritative reference
IAB — Full-Funnel Measurementiab.comSee also
Reference docs
Building against Trillboards?
Our developer reference covers the DSP API, partner SDK, proof-of-play verification, and the sensing pipeline that powers buyer-grade audience signals.
View developer docs