How DOOH impressions are calculated
Learn how DOOH ad plays become estimated audience impressions, how CPM is calculated, and what reach and frequency add to a campaign report.
A digital billboard can put a brand in front of people on their way to work, shopping or heading home. Unlike an ad delivered to one phone, a single play on that screen may reach several people at once. Digital out-of-home advertising uses audience estimates to account for that shared exposure, which is why its impression reports need a little explanation.
For example, 1,000 ad plays with an average estimate of 12 audience exposures per play produces 12,000 estimated impressions. It does not establish 12,000 different people or 12,000 people who noticed the ad.
Measurement is also an active industry issue. In March 2026, Geopath and OAAA announced an Ipsos pilot to evaluate updated audience measurement methods. When you compare quotes, ask how each seller calculates the audience in your proposal.
Separate ad plays from audience impressions
An ad play is the supplier’s recorded display event. Ask whether that record comes from a screen playback log or another delivery signal, and whether any independent verification is available. Audience impressions estimate the exposures associated with those plays under the supplier’s methodology. Neither number is a count of customers who remember the message.
Google’s DOOH buying documentation explains that publishers use an impression multiplier to estimate viewers per ad. It also distinguishes publisher impressions from served impressions in Display & Video 360. Use the DOOH-specific reporting fields when you evaluate those campaigns.
If your report contains only the audience total, you cannot reconstruct delivery. The play count, measurement provider and applicable multipliers help you reconcile the estimate. They do not independently verify that the screen displayed the ad or that someone saw it.
Understand the multiplier
The 1,000-play example assumes a constant multiplier of 12. That is illustrative arithmetic, not a benchmark for a roadside screen or venue.
Now split those plays: 600 at a multiplier of 15 and 400 at a multiplier of 6. The first group contributes 9,000 impressions; the second contributes 2,400. The total is 11,400. Applying 12 to every play would instead report 12,000, so the time and screen breakdown matters.
For a buy reported using per-play multipliers, estimated audience impressions = the sum of ad plays × the applicable audience multiplier for each screen and reporting period. Confirm whether your seller uses this calculation or another audience model before applying it to a quote. A campaign-wide average needs to be weighted to the plays in those groups, rather than selected because it produces a convenient total.
Who is included in the audience estimate?
A venue’s visitor total is not enough. People may enter through another door, face away from the display or visit outside the ad schedule. The audience method should explain which viewing area and times contribute to the estimate.
OAAA’s May 2021 impression measurement guidelines recommend an opportunity-to-see measure with documented viewsheds, transparent data sources, and time-based reporting. A viewshed describes the area from which a display can be seen. Some methodologies further qualify exposures for the likelihood of seeing the advertising.
Two sellers can use the word “impressions” for differently qualified exposures. Check how each seller defines an exposure before comparing their rates.
For a December mall buy, look at the dates behind the audience forecast. An estimate based on ordinary weekday traffic needs an explanation before it is used to represent holiday weekends. The venue name alone does not establish seasonal delivery.
Calculate cost per thousand impressions (CPM)
At a $120 media cost, the original 12,000-impression example has a $10 audience CPM: $120 ÷ 12,000 × 1,000.
Use 1,000 ad plays as the denominator and the result is $120 per thousand plays. The twelvefold difference comes from changing the unit, not from changing the price.
If one quote includes $120 in media and another adds $30 in fees, show the extra charge. The second total is $150, or $12.50 per thousand of the same 12,000 estimated impressions. Compare media CPMs separately from the full advertiser cost. These examples assume the cost is priced against the stated audience impressions; a fixed-duration or share-of-loop quote needs its own delivery estimate before you calculate an equivalent CPM.
The comparison also needs the audience and placement context. A lower CPM against a broad audience does not establish that a screen is better for a business trying to reach customers in a specific service area.
Add reach and frequency where supported
Impressions include repeated exposures. Reach estimates distinct people, while frequency describes the average exposure among the reached audience within the reporting period. Geopath’s reach and frequency primer explains this distinction.
If the report estimates 3,000 distinct people for those 12,000 impressions using the same audience definition, methodology and reporting period, average frequency is four. Without a reach estimate, you cannot infer 3,000 people from the impression total alone.
Reach summed across individual screens can count the same commuter several times. A campaign-level estimate needs to account for overlap. Without that step, the screen totals cannot support a claim about distinct people reached.
From forecast to delivered campaign
Your proposal from a buying partner should link the screen list and flight dates to the promised reporting detail. If results are available only for the whole package, you cannot later expect a screen-by-screen evaluation.
Suppose a screen misses two days. Ask how the seller reconciles the missed plays and audience estimate, and whether replacement delivery occurred elsewhere or later. A makegood may restore a contracted total without recreating the original location or dates. Keep planned delivery, reported delivery and any makegood separate so you can see what ran.
The impression report explains the estimated audience exposure. It cannot, by itself, establish incremental foot traffic, attention, or revenue.
For the broader inventory options, read Floodlight’s DOOH advertising guide. The measurement method should be part of any quote for those placements.
Frequently asked questions
Does one billboard ad play equal one impression?
One play can generate several estimated audience impressions. The relationship depends on the measurement method and audience estimate for that screen and time period.
Do DOOH impressions prove someone looked at the ad?
Ask which exposure definition is used. Opportunity-to-see and likelihood-to-see estimates describe different measurement approaches. Neither proves an individual remembered the creative or became a customer.
Are impression estimates the same across all DOOH sellers?
No. Request the provider, method, data dates, and reporting definition before comparing estimates. Consistent labels alone do not establish consistent measurement.
Work with Floodlight
Floodlight helps brands build and manage campaigns across connected TV, streaming audio and podcasts, digital out-of-home, display advertising, and ChatGPT advertising. We also provide AI research and visibility and programmatic media planning and buying. Talk with us about evaluating DOOH inventory and building a media plan around your audience, goals, and measurement needs.