Matching Algorithm

Matching is the central component of Adlocaite: it decides which advertisement can play on which screen, at which moment, and at which price.

Two phases: matching and delivery

Adlocaite separates the decision from the playout:

Continuous matching runs in the background, at regular intervals. It works through a continuously updated database of screens and advertisements and computes, for every screen, which ads are playable there and at what bid.
Delivery happens when a publisher's screen requests an advertisement. At that moment nothing needs to be computed anymore: the precomputed match for this screen and moment is handed out as an offer.

This separation is why offer responses are fast, and why every delivered spot is backed by a full targeting evaluation rather than a split-second guess.

Continuous matching

The platform maintains two continuously updated inventories: the available screens with their data-driven attributes (see Data Sources), and the active advertisements with each advertiser's targeting, which is configured manually, refined together with our AI, or delegated to it entirely.

At regular intervals, the matching computes playable ads for every screen. Targeting values play two different roles in this:

Fixed values are hard constraints. Hours, weekdays, or holiday status either match or they do not: an ad targeted at evening hours is simply not playable on a screen slot at noon.
Relative values are scored. Demographics, income, weather, or intent-based audience fit are a matter of degree: the better the screen's expected audience fits the ad's targeting, the higher the match scores.

Targeting criteria

Targeting consists of different values that are considered in the matching decision, including both fixed and relative targeting parameters.

Fixed values

These targeting parameters act as hard constraints:

Time-based targeting

Target hours (0-23)
Target days of the week (Monday through Sunday)
Target holiday status (on holidays, outside holidays)

Location-based targeting

Target Points of Interest (POI) such as airport, train station, university, business district
Target Points of Sale (POS) such as fashion and luxury, electronics, supermarket, restaurant

Relative targeting values

These parameters are scored by degree of fit:

Demographics

Target age groups, from teens to seniors
Target gender
Target income brackets

Environmental factors

Target weather forecast (positive, negative)
Geographic targeting via regions

Bid calculation

For every match, the platform also computes a bid for that screen. The bid derives from two inputs: the target CPM the advertiser has configured, and how well the screen fits the targeting. A screen whose expected audience matches the campaign closely is worth more of the advertiser's budget than a loose fit, and the bid reflects that.

The result of continuous matching is therefore not just a list of playable ads per screen, but a priced one: every playable ad carries the bid it would pay for a spot on this screen.

Delivery

When a publisher's screen requests an advertisement (see Offers), the platform hands out a precomputed match as an offer with a concrete bid price. When several ads are playable on a screen, delivery rotates among them, so the screen shows changing creatives over time: more visual variety means more attention per spot. The publisher's system accepts or rejects the offer; acceptance creates the deal and the spot plays.

Delivery is deliberately simple. The intelligence sits in the continuous matching that happened before the request ever arrived.

Optimization priorities

The matching system prioritizes screens with optimal targeting over other metrics such as raw impression volume.

Targeting-first approach

The system ensures that:

Optimal targeting takes precedence: Screens with the best targeting match are preferred over those with higher reach but lower targeting accuracy
Audience-focused delivery: Advertisement is always displayed to the target audience, regardless of audience size
Quality over quantity: Targeting accuracy is prioritized over reach metrics and other performance indicators

Result

This approach ensures that advertisement is always delivered to the target audience, independent of audience size, with optimal targeting being the primary consideration over raw impression or audience numbers.