What ran

Ad delivery, hourly

Completed plays per screen per hour, counted the way billing counts them, with distinct creatives and ads, seconds played, plays that passed invalid-traffic filtering, average CPM, gross revenue and the demand source that filled the most plays.

Request a sample All datasets

The question it answers

Which ads ran on which screens, when, and what did they earn?

532,646

plays

4,610

screens

97,234

screen-hours with plays

Measured over the 28 days to October 11, 2026.

The dataset

Grain: Hourly.
Cadence: Hourly rows, one file per day.
History: From April 2, 2026.
Format: Parquet or JSONL.
Delivered to: Your S3 bucket, your GCS bucket, or a time-limited download link, with a manifest and a schema.
Method: How each measure is produced and how to add rows up: the measurement methodology.
On the network: Demand ecosystem.

What buyers do with it

Fill, CPM and revenue benchmarks for screen media by market and hour.
The audience of every hour that ran ads: join place activity on screen and hour.
Supply-side research on programmatic out-of-home demand.

The table

Read from the export registry each time this site is built.

`ad_delivery_hourly`

Hourly, from April 2, 2026

Grain: One row per hour_start and screen_id.
Delivery: One file per day, as Parquet or JSONL.
Basis: Measured.
Definition: Version 3.

Column	Type	What it holds
`hour_start`	timestamp	Start of the UTC hour. A play counts in the hour its kept report completed.
`screen_id`	string	Trillboards screen id (24-character hex); joins dim_screen.screen_id.
`completed_plays`	integer	Completed plays as our billing policy identifies them: the play key on one screen, each play once, on its first day. Unit: plays.
`distinct_creatives`	integer	Distinct creative ids among the plays (the IMA house creative placeholder is not a creative). Unit: creatives.
`distinct_ads`	integer	Distinct ad ids among the plays (the pod placeholder is not an ad). Unit: ads.
`played_seconds`	double	Sum over the plays of each play's longest positive duration over every report of the play that day. Unit: seconds. Null: No play of the hour carried a positive duration on any of its reports.
`played_duration_plays`	integer	Plays that carried a positive duration: the denominator of a mean play length (played_seconds / played_duration_plays). Unit: plays.
`givt_passed_plays`	integer	Plays that passed general invalid-traffic checks. Unit: plays.
`avg_cpm`	double	Mean CPM of the plays, USD. Unit: usd_per_1000_plays. Null: No play of the hour carried a price.
`gross_revenue`	double	Sum of the plays' gross revenue, USD. Unit: usd. Null: No play of the hour carried a price.
`modal_source_provider`	string	Most frequent demand source of the plays by exact count, ties to the alphabetically first. Null: No play carried a real provider.

Request a sample

Tell us the markets and the weeks you want. The sample is cut from production by the same export that makes the full delivery.

Request a sample

All datasets · Next: Screen master