Display and CTV buyers can choose from mature invalid-traffic detection, viewability standards, and pre-bid avoidance built over more than a decade of fraud fights. Programmatic audio and digital out-of-home (DOOH) buy through the same pipes, the same DSPs, and often the same OpenRTB bid requests, but the measurement layer behind them is thinner. In audio, most signals browser-based verification relies on do not exist. In DOOH, there is no individual viewer to verify at all, only a screen, a playout log, and a statistical estimate of who might have seen it. That gap is where budget is least protected, and it is worth understanding before the next plan shifts spend into those channels.

Why Audio Is Hard to Verify

Display verification rests on a tag running in a browser or app. It can inspect the environment, measure viewability, and look for signs of automation. Much digital audio, especially podcasts and streaming radio, is delivered through server-side ad insertion (SSAI). The ad is stitched into the audio stream on a server, and the listener’s device may never run a verification script. What the buyer sees is a server reporting that it delivered an ad to a device identified by an IP address and user agent it passes along.

That structure is exactly what fraudsters target. In February 2023, DoubleVerify disclosed BeatSting, which it described as the first large-scale ad impression fraud scheme aimed at digital audio. The operators spoofed residential IP addresses and audio apps and ran fake SSAI servers that generated audio ad requests which looked legitimate to buyers. DoubleVerify estimated the scheme cost unprotected advertisers up to $1 million a month and traced the fraud family back to 2019. It later reported a second audio scheme, FM Scam.

The lesson from both is structural. When the server making the ad request is also the source of the delivery data, the buyer is trusting the server’s account of events. Detection has to rely on pattern analysis across traffic, such as impossible listening volumes, mismatched device signatures, and IP ranges that do not look like homes. It cannot rely on direct observation of the listener.

What Audio Verification Covers Today

Verification vendors have expanded into audio, but coverage is uneven by type of protection. The clearest recent example: on June 11, 2026, Integral Ad Science extended its Total Media Quality measurement to YouTube Audio Ads campaigns. That expansion is for brand safety and suitability, confirming the ad runs next to acceptable content. IAS has also worked on episode-level pre-bid contextual controls for podcast inventory through The Trade Desk. Both are useful. Neither is the same as invalid-traffic detection.

That distinction is the practical gap. For audio, brand-safety and contextual coverage has been extended faster than fraud and IVT measurement, because content classification works on the audio or transcript itself while fraud detection has to work around SSAI’s blind spots. IAS’s audio offering does describe invalid-traffic detection, audibility, and quartile completion metrics. Buyers should confirm, per publisher and per delivery path, which of those are measured directly and which are inferred from server logs.

Why DOOH Is Hard in a Different Way

DOOH does not have a user device problem. It has no user at all. A screen in a transit station or a mall plays an ad, and the number of people who saw it is an estimate. OpenRTB 2.6 gives this a formal home: the dooh object describes the venue and screen, and the qty object carries an impression multiplier with fields for the multiplier value, its source type, and the vendor that produced it. A DSP bids a CPM for one play, and the exchange multiplies by the modelled audience to get the billable figure. Fractional multipliers are common, because the number is a statistical model, not a head count.

This creates fraud and quality risks that display verification tools were never designed to catch:

  • Playout fraud. Was the ad actually shown, on the screen claimed, at the time claimed, for the full duration? That is a proof-of-play question, answered by the media owner’s player logs and, where available, third-party photo or sensor checks.
  • Multiplier inflation. If the audience multiplier is too high, every impression is overbilled even when every play is real. The multiplier’s source and the vendor’s method matter as much as the playout.
  • Screen and venue misrepresentation. A bid request describing a premium venue type that is actually a lower-traffic location is the DOOH equivalent of domain spoofing.

The IAB’s DOOH Measurement Guide, published in July 2025, set out definitions for opportunity-to-see and likelihood-to-see, dwell time, and attribution methods, and stressed accredited methodologies and third-party verification. In practice, DOOH verification is largely a specialist market of proof-of-play and audience-validation providers, not an extension of the display verification stack. MRC accreditation is also product-specific: a provider accredited for one format or channel is not automatically accredited for DOOH.

What Buyers Can Actually Check

The gap does not mean these channels should be avoided. It means the verification plan has to be built per channel rather than copied from display.

For audio:

  • Ask whether delivery is client-side or SSAI for each publisher and path. SSAI traffic should be covered by a vendor that analyses server-side patterns, not only by a tag.
  • Separate brand-safety coverage from IVT coverage in the verification contract. Having one does not mean you have the other.
  • Buy through the shortest supply path available. Reselling layers add more servers whose accounts of delivery you have to trust.

For DOOH:

  • Require the qty multiplier, its source type, and its vendor in reporting, not just the billed total.
  • Ask for proof-of-play data, including player logs at minimum and third-party validation where the budget justifies it.
  • Check whether the audience measurement provider is accredited for DOOH specifically, not just for another channel.
  • Compare multipliers for the same venue types across media owners. Large, unexplained differences deserve an explanation.

Audio and DOOH are growing in programmatic plans partly because they look clean compared with the open web. Some of that cleanliness is real, since screens and podcast apps are harder to fake at scale than websites. Some of it reflects how little has been measured. Buyers who ask which protections apply to which delivery path, and who treat multipliers and server logs as claims to verify, will know which of the two they are paying for.