How to Audit an Impressions Claim Before You Quote a CPM

Before you turn an impressions figure into a price, you should be able to take it apart. Learning how to audit an impressions claim protects you in both directions: it stops you overpaying for inventory you buy or resell, and it stops a buyer's analyst catching a hole in a number you published. This post gives you a five-question audit and shows how a verified count becomes a defensible CPM.

It is the practical follow-up to why you should never sell on an unverified number, and it sits under the OOH audience measurement guide.

Key takeaways

• Audit every impressions claim with five questions: period, method, zone, error range, verifiability.
• "How many impressions does a billboard get?" has no universal answer — it depends on that face's measured traffic and the multipliers applied.
• CPM is price ÷ impressions × 1,000, so an inflated impressions figure quietly inflates a CPM a buyer will renegotiate.
• A verified count gives you a CPM you can hold in a negotiation instead of defending a guess.

How to audit an impressions claim: the five-question test

Run any impressions number — yours or someone else's — through these five questions. A confident answer to all five is a number you can price against. A gap in any one is a discount waiting to happen.

1. Over what period was it measured?

An impressions figure with no window is meaningless. Ask for the exact days and hours. A number from a single peak afternoon projected onto a whole month is a very different thing from one sampled across dayparts and weekdays. If the answer is "typical" or "annual average" with no observed window behind it, treat it as modelled, not measured.

2. By what method?

How was the audience actually counted? Video counts, sensors, mobile-panel modelling, and a person with a clicker are not equivalent, and each has known blind spots. If the method can't be named in a sentence, the number is a claim, not evidence. The difference between panel estimates and direct counting is spelled out in panel data vs ground-truth counting.

3. In what visibility zone?

An honest OOH number counts only the audience with an opportunity to see the face — the traffic inside its visibility zone — not every vehicle on the whole road. Ask what was counted. A figure built from total road traffic rather than the zone is inflated by everyone who never had a chance to see the ad.

4. What is the error range?

Every real measurement has uncertainty. Ask for the range and the confidence behind the headline. A number offered as a single exact figure with no margin is either naive or hiding something. A range like "34,000–50,000, high confidence" is a sign of a number produced honestly — see what "accurate" really means for an impressions number.

5. Can it be independently verified?

Finally: can a third party check it without trusting the seller? A method note, a proof clip, and a QR verification page are what separate a certificate from a spreadsheet cell. If nothing can be verified, price as if the number were unproven — because it is.

"How many impressions does a billboard get?"

This is the question buyers and search engines ask, and the honest answer is: it depends entirely on the face. Two hoardings a mile apart can differ several-fold. The number is a function of the traffic in that face's visibility zone, the dayparts it faces, and the occupancy and visibility multipliers applied. Anyone quoting a single figure for "a billboard" in the abstract is guessing. The only reliable answer is to measure the specific face — which is exactly what an Audience Certificate does.

From a verified count to a CPM

Once a number passes the audit, turning it into a price is simple arithmetic:

CPM = (price ÷ impressions) × 1,000.

So if you want to charge $500 for a four-week flight on a face verified at 168,000 impressions over that period:

• CPM = (500 ÷ 168,000) × 1,000 ≈ $2.98.

The impressions figure is the denominator of your price. That is why an inflated number is self-defeating: pad the impressions and your CPM looks artificially cheap, so you leave money on the table; a buyer who then re-measures finds fewer impressions and demands the CPM back. A verified count lets you quote a CPM you can hold — and defend line by line if the buyer runs their own audit.

Red flags that should stop a quote

Some signals should make you pause before you price against a number at all:

• A round headline with no range. "500,000 a month" and nothing else almost always means nobody measured it.
• "Traffic" used where "visibility zone" belongs. If the figure is built from total road counts, it is inflated by everyone who never had a chance to see the face.
• An annual figure with no observed window. A yearly average with nothing behind it is a model, not a measurement.
• No proof clip and no verification link. If nothing can be shown or scanned, there is nothing to check.
• Accuracy claimed as a single badge. "98% accurate" with no error range hides more than it reveals — see what "accurate" really means.

Any one of these is reason enough to treat the number as unproven and price accordingly.

Audit your own numbers first

The five questions are not just for inventory you buy. Run them on your own faces before you publish a rate card. If you can't answer all five for your own numbers, a buyer's analyst will find the gap, and you would rather find it first. Auditing yourself is also the fastest way to spot which faces genuinely justify a premium and which have been coasting on an optimistic guess.

Make every number audit-proof

The simplest way to pass your own audit is to measure. Certify one face for $99 to get a number that answers all five questions out of the box, or see pricing to do it across a network. Then go deeper on the trust question in what "accurate" really means for an impressions number, or return to the measurement guide.