Programmatic needs the venue's numbers
Programmatic DOOH promised digital-style buying for physical screens, and the machinery exists: DSPs bid, SSPs offer, deals clear in milliseconds. What the machinery runs on is data about the physical world, and that data has to come from the venue. A venue with measured footfall has three distinct products to feed into the pipes; a venue without one contributes a guess and takes whatever price the guess earns.
How does footfall data reach a DSP?
Three routes, in increasing sophistication. Audience packs: per-screen, per-daypart audience figures attached to the inventory an SSP offers, refreshed on a schedule, which set baseline pricing. Triggers: near-real-time conditions (current traffic above a threshold, queue forming, zone busy) exposed via API so buys activate when the audience is actually there. Reporting feeds: post-campaign delivery data at the same grain as the buy, so the DSP's reporting reconciles with the venue's measurement. Most venues start with packs and grow toward triggers.
Audience packs: the baseline product
The pack is a structured answer to "who is here, when": passing counts and dwell per screen position, split by daypart and day of week, refreshed frequently enough that seasonality shows up. Two properties make a pack sellable. It has to come from measurement rather than a licensed model (buyers discount modeled audiences, and the difference is covered in how DOOH audiences are measured), and it has to be aggregate by construction: counts and averages, never anything resembling a per-person record.
Triggers: selling the moment
The programmatic-native product is conditional inventory: this screen, when the corridor is running hot; the food-court screens, when lunch queues form. Triggers need a live data path from the counting layer to an API the SSP or DSP can poll or subscribe to, with latency measured in minutes, not days. They monetize precisely what makes physical venues valuable, presence you can time, and they are the natural continuation of the attribution work described in programmatic DOOH attribution.
The privacy line that must hold
Digital ad tech's instinct is per-user data, and every integration conversation eventually drifts toward it: device IDs, retargeting pools, exposure graphs. The venue's answer has to be structural, not contractual. Ariadne's measurement carries no identifiers at any point (no MAC addresses, no device IDs, no images), so the feeds built on it are aggregate because the underlying data is aggregate; there is nothing per-person to leak, sell, or subpoena. That property, argued in full in DOOH without facial recognition, is what lets a European venue participate in programmatic without importing digital's consent problem.
Build order for a venue
Measurement first, at every screen position, continuously. Packs second, once a full season of baseline exists. Triggers third, where the venue's traffic actually varies enough to sell moments. Reporting feeds throughout, at the buy's grain. The measurement standards these feeds answer to are collected in DOOH measurement standards.
