Demographic targeting in OOH rests on an indexing methodology, and understanding it (including its limits) is what separates confident buying from hopeful buying.

Audience profiles per unit are built from anonymized movement patterns combined with demographic data from census sources and established third-party providers such as TransUnion, LiveRamp, and Claritas, describing the population that actually passes each unit; not just who lives nearby, but who travels through.

An index compares a unit's audience concentration for a segment against the general population baseline, where 100 is average: an index of 150 for "household income $150K+" means that audience passes this unit 50% more than average. Indexes make units comparable on audience fit the way CPM makes them comparable on cost; use both together.

Three practical rules: a high index on a low-traffic unit can mean fewer target impressions than an average index on a busy unit (index times volume is what you're buying); indexes are strongest for broad, movement-correlated segments (income, age bands, commuter patterns) and softer for narrow interest segments; and clusters of units with consistent indexes are more trustworthy than one outlier unit.

This is probabilistic audience description of physical spaces, not individual-level targeting; nobody is verifying the age of each passerby. It reliably tilts a plan toward your audience; it cannot promise each impression lands on your ICP. For plans needing tighter fidelity, layering first-party data (see the audience targeting article) is the strongest available signal.

Describe your ICP concretely (roles, behaviors, places they reliably go) and planning shifts from demographic indexes to venue and movement logic: the office districts, gyms, airports, and routes where that person predictably appears. For narrow ICPs, venue logic usually outperforms demographic filters.