DOOH ad verification
The screen owner’s guide to DOOH ad verification
If one of your screens went dark at 2pm yesterday, would tomorrow’s report say so? For most networks the honest answer is no. This guide shows what your proof really covers, what the gap costs you, and what to change.
Updated 31 October 20269 min read
The short answer
Your report says played. Does the screen agree?
An advertiser buys a slot expecting their ad to run, in full, on a screen that was on and visible. What most operators hand back is a log that says the ad was sent. The gap between “dispatched” and “displayed” is where DOOH revenue leaks away.
It isn’t a fault in your setup. It’s how digital signage is built: the CMS knows what it told the player to do, and the player knows what it tried to do, but neither can see what reached the glass. A frozen frame, a loose cable or a panel that rebooted mid-loop all look the same as “played successfully” from inside the system reporting on itself.
So the question isn’t whether your log is accurate. It’s what your log can’t see.
Check your own proof
Six things that go wrong on real screens every week. Tick what you rely on today and see which of them a buyer would never hear about.
5 of 6 would still go in your report as delivered.
- The screen was off or black Missed
- A frozen frame held the whole slot Missed
- No signal, or the wrong input on the panel Missed
- Something was blocking the panel Missed
- The wrong content was showing Missed
- The content stopped before it finished Caught
Based on what each method can and can’t see, explained in the methods compared.
What the gap is costing you
Buyers price risk into a rate whether or not you do. Every “missed” above is a reason for them to.
You give inventory away to be safe
Networks routinely run a cushion above the sold impressions so the final count can't fall short. That cushion is inventory you never sell, given away on every campaign, because the number can't be proven precisely.
Downtime you hear about from the client
A screen goes black or freezes and nobody notices until the advertiser mentions it, after the campaign has already run under-delivered. Then it's a make-good, and a harder conversation at renewal.
Premium sites sold at unverified rates
Without evidence, buyers price in the chance a slot didn't deliver, even on your best sites. Unproven inventory sells like commodity inventory, whatever it cost you to build.
The other side of each of these is where independent proof pays: sites with verified delivery can hold a premium over unverified sites in the same category, a disputed invoice becomes a lookup rather than a judgement call, and a track record of verified delivery is what moves an advertiser from spot bookings to a longer contract.
The four ways to prove a play, compared honestly
“DOOH verification” covers genuinely different techniques. None of them is fraudulent or useless; they answer different questions. What decides their weight is how far each one sits from the system it’s checking.
Method
How strong
Scheduling log
WeakThe spot was booked and dispatched.
Who says so: You do.
Blind to: Cannot see the screen at all.
Player playback log
WeakThe device tried to play the file.
Who says so: Your player does.
Blind to: Reports on itself. A frozen frame still logs as played.
In-creative measurement
SoundThe creative loaded and ran where it was placed.
Who says so: The buyer’s own code does.
Blind to: Rides in the delivery path it is reporting on.
Independent sensor
DefinitiveContent appeared on the glass, matched to the campaign.
Who says so: Something with no stake in the answer.
Blind to: Needs hardware on the screen.
The lower rungs are reliable at their own question, a scheduling log is a fine record of what you booked, and a playback log is good fleet telemetry. They are weak as evidence that an ad appeared for one reason only: the answer to “who says so” is the same party being asked. Read the column and the ladder explains itself.
Player playback logs
The player records what it opened, when and for how long, and reports that back to the CMS.
- Good for
- Free, on every player, with exact start and stop times. Good for fleet health: crashes, errors, connectivity gaps.
- Misses
- It measures what the player meant to do, not the screen. It can't see a black display, a stuck frame or a disconnected cable, and it's the system grading its own homework, so agencies increasingly discount it on its own.
Screenshot sampling
The player, or an agent on it, grabs a still of its output every so often as a spot-check.
- Good for
- Looks at the output, not just the software's claim. Useful for creative QA, and cheap to store.
- Misses
- It's sampled, so a freeze or a drop-out between snapshots goes unseen, and it comes from the same device being checked. It can't confirm a play ran in full.
In-creative measurement
Measurement code ships inside the creative, usually as an HTML5 tag, and reports from inside the player's render on any network the ad runs on.
- Good for
- Belongs to the buy side, travels with the creative across networks with nothing for operators to install, and gives one campaign view instead of a log per operator.
- Misses
- It runs inside the render, so like a log it can't see the physical screen: panel power, signal, wrong input or something blocking the view. It needs an HTML5 unit to ride in.
Independent sensor
A separate device watches the screen continuously and fingerprints what actually appears on it.
- Good for
- Independent of the playback system and continuous: catches black screens, freezes, wrong content and downtime as they happen, and produces matched, timestamped evidence a buyer can audit.
- Misses
- It's hardware at the screen, so a cost per site, and it's only as good as its matching: it has to tie what it saw to the right campaign.
What buyers accept, and when it stops being enough
Most agencies still take a playback log for routine billing. The line moves fast when there’s a dispute, an audit or a premium rate on the table: that’s when independent evidence stops being a nice-to-have and becomes the thing that settles the invoice.
| Evidence | Accepted for | Rarely enough on its own for |
|---|---|---|
| Scheduling log | Operational reporting, billing reconciliation | Delivery disputes, make-good negotiations |
| Player playback log | Fleet health, uptime monitoring | Independent proof a specific spot displayed |
| Screenshot sampling | Spot-checks, creative QA | Continuous delivery guarantees |
| Independent sensor | Advertiser-grade reporting, contested invoices, premium pricing | Nothing on this list |
How a verified play is made
Whatever the method, a play becomes evidence the same way:
- 1
Log what was scheduled. Record what content was pushed to which screen, and when, from your CMS. This is the baseline every method below is checked against.
- 2
Capture what the player reports. Collect the player's local playback log: what it actually opened, for how long, and any errors it hit. This is evidence of intent, not proof of display.
- 3
Sample or continuously observe the physical screen. Use screenshot sampling or an independent sensor to check what was actually showing on the screen, separately from what the player claims it played.
- 4
Match observed content back to the campaign. Fingerprint or otherwise identify the observed content and match it to the scheduled spot, so a verified play can be tied to a specific campaign and invoice line.
- 5
Package it as evidence. Turn matched, verified plays into a report an advertiser or agency can audit, not just a delivery percentage, but a defensible record.
Three things to do this week
- 1
Test one campaign against the glass
Pick a campaign and a site you can see, in person, on camera or through photos. Put what your report says next to what the screen actually showed. One mismatch tells you more than any guide.
- 2
Ask your biggest buyer what they'd accept
Not for routine billing: for a disputed spot at full rate. Their answer is the bar your evidence has to clear, and it's better to hear it now than in a dispute.
- 3
List the screens where a gap costs most
Your premium sites, the ones buyers query, and the ones that are expensive to visit. That's where independent proof pays for itself first. Start there, not everywhere.
Questions screen owners ask
- What evidence do advertisers accept for DOOH delivery?
- For routine billing, most advertisers still accept a delivery report built from playback logs. The bar rises the moment delivery is questioned or a premium rate is on the table: in a dispute, a self-reported log carries little weight, and agencies increasingly ask for independently observed evidence, such as sensor-based verification, before they will settle at full value.
- Can I verify my screens without replacing my player or CMS?
- Yes. Independent sensor verification attaches alongside whatever is already driving the screen. Vojo's Harvester, for example, works with any player or signage system: it watches the screen itself, so nothing in your existing playback stack needs to change.
- What does DOOH verification cost?
- Playback logs are effectively free, since every player produces them. Independent sensor verification adds a device per screen plus a licence. On Vojo the verification portal itself is free, and the rest is priced in your quote: either a monthly licence per Harvester, or pay as you go on what you verify, depending on which suits your network. The licence is a fraction of the premium a verified site can defend, so operators typically verify their premium sites first and expand as the verified inventory earns higher rates.
- Does verification actually change what inventory earns?
- Verified delivery changes the negotiation. Operators can defend rate cards with evidence instead of estimates, disputes settle on data rather than goodwill, and make-goods stop being the default outcome of every query. The screens that benefit most are premium sites currently selling at unverified rates.
Patented Proof of Display
Where Vojo fits
Vojo Passport™ rides inside the creative on every screen it plays on, so the campaign carries its own check-in from inside the render: which creative ran, where, and whether it finished. Nothing for the operator to install.
On the screens where proof matters most, a Vojo Harvester™ adds the independent sensor described above. It watches the glass, fingerprints what genuinely appeared and matches it to the same play, so a black screen, a frozen frame or the wrong input is caught instead of billed. Harvesters always ship with Passport.
Every play gets a vTrust Score from the evidence it actually carried. The Proof of Display method is patented in the UK (GB2583366, GB2597434) and the US (US 12,176,999), with a European patent pending. For how the pieces fit together, see how Vojo works end to end.
The DOOH verification glossary defines every term used here, and Proof of Play vs Proof of Display goes deeper on the difference.
Vojo's Proof-of-Display technology is protected by granted patents in the United Kingdom (GB2583366, GB2597434) and the United States (US 12,176,999), with a European patent application pending.