CPM vs per play, forecast versus fact.

What CPM and per play each mean

CPM stands for cost per mille, the cost of one thousand impressions. An impression is one estimated view of your ad, so a CPM figure is really a price on a forecast: a network models how many people are likely to pass a screen while your ad runs, and charges for those modelled thousands. CPM is the long-standing currency of reach buying, and typical managed digital out-of-home sits around $3 to $15 CPM depending on market and format.

Per play is the cost of one play, and a play is a single appearance of your ad on one screen. It is not a forecast of who saw it; it is a record that it ran, logged with the screen, the timestamp and the location. On Blindspot the per-play price is shown on every screen card before you book, from about $0.23 a play in an urban market up to premium landmark rates, and that is the unit you are billed on. A play is the thing you buy; the price is set against it.

Forecast versus fact

The heart of the difference is what each unit prices. CPM prices something that has not happened yet, an estimate of reach. Per play prices something that did happen, a verified display. That changes what you can hold the campaign to: a CPM buy is measured against the forecast it was sold on, while a per-play buy is measured against a log of real appearances, the same record that attribution reads to tie a screen to a store visit or a web session.

The CPM range reflects typical published managed digital out-of-home pricing and can change by market and format. It is shown here only as the industry comparison unit; Blindspot itself prices per play, never on CPM. See real per-play prices by city and format in the per-play pricing index.

Why per play is more efficient

Two things make per play the more efficient buy. First, accountability: because you pay for real displays, the spend maps to something that happened. There is no gap between the forecast you were sold and the delivery you got, which is why a per-play log is the receipt behind measurement. Second, control: buying plays lets you decide exactly when they run. You can schedule each screen by the hour and drop the empty overnight and dead midday windows a fixed flight pays for, which typically removes 30% or more of the cost. The freed budget then buys more plays in the windows that convert.

That efficiency holds at any budget. A first campaign of a few hundred dollars spends every play on the hours its audience is out, and a worldwide flight does the same across thousands of screens, so a budget buys the real exposure it needs rather than filler plays. There is no minimum spend, and Blinky, the free AI planner, will read a one-line brief and propose the plays and hours that carry your audience.

The gain shows up in outcomes, not just in the price. Because a play is a verified event, Blindspot can tie the screens that ran to real actions and report the cost of each: about $0.82 per incremental store visit, $0.80 per incremental web visit, and $5.75 per incremental online purchase, against typical paid-social costs of $15 to $40 per acquisition. Those figures exist because the underlying unit is a logged play, not a modelled thousand; you cannot audit a forecast the same way. See how it is measured in the attribution guide.

None of this rules CPM out entirely. If you are buying broad reach and the forecast itself is the currency you report on, CPM still fits. But for buyers who want to know their money bought a real appearance, and who want to cut the hours nobody is watching, per play is the more efficient and more honest unit. Compare the field in the DOOH platforms guide, or see the exact prices in the billboard cost guide.