A campaign can report millions of impressions and still leave one question unanswered: how many reflected a real opportunity for a person to notice the ad?
Impression counts remain useful. They describe scale and delivery. The problem begins when they are treated as proof of human attention, or when fraudulent activity makes the underlying audience unreliable.
Better measurement starts by separating four ideas that are often compressed into one number.
1. Was the delivery valid?
Before asking whether an ad earned attention, advertisers need confidence that the impression came from legitimate inventory and a genuine device or user.
AppsFlyer’s State of Fraud for Marketers 2026 analysed 106.4 billion installs across 246,000 apps. It found that organic traffic accounted for 52% of fraudulent installs, while owned-media fraud rose 221% year on year.
The lesson is uncomfortable but useful: a familiar channel is not automatically a clean channel. Validation, anomaly detection and transparent supply paths are the foundation of every metric that follows.
2. Did the ad have an opportunity to be noticed?
For visual media, viewability asks whether enough of an ad was on screen for long enough to create an opportunity to see it. For audio, the equivalent question includes whether sound was genuinely available to hear.
Audiomob’s patented volume detection and audio passback technologies are designed to address this. These technologies help distinguish basic playback from delivery under audible conditions. Our Audibility blog outlines how this works.
Opportunity is still not attention. It is the qualifying condition that makes attention possible.
3. What evidence suggests attention?
The IAB attention measurement framework, developed with MRC and CIMM, makes a crucial distinction: viewability establishes the opportunity to be seen while attention measurement examines the likelihood that an ad was noticed or engaged with.
Attention measurement can draw on signals such as exposure duration, interaction and audio completion. Outcome measures can then assess whether exposure was associated with recall, consideration or action.Each method has limits, so transparency about what was measured matters as much as the final score.
For audio, listen-through rate shows whether the ad played to completion. Audibility signals provide additional evidence that sound was available while brand recall research assesses whether the message was remembered.
4. Did attention create an outcome?
Attention is valuable because of what it can enable, not because it creates another dashboard metric.
The campaign recorded a 95.2% listen-through rate. Post-campaign research reported 27% recall across the full audience while nine in ten respondents said they had taken action or intended to act after hearing the ad. The full HMRC case study connects delivery with a real communications objective.
Standardisation supports confidence
Independent verification helps advertisers compare performance across environments. Audiomob is compliant with the IAB Tech Lab’s Open Measurement SDK, which provides a standardised route for third-party measurement in apps. Audiomob’s brand-safety overview explains how OM SDK compliance fits alongside inventory controls and fraud prevention.
The right measurement ladder is simple to state, even if it is demanding to implement: validate the impression, confirm an opportunity to notice, examine evidence of attention and connect it to an outcome.
That is how advertisers move from counting activity to understanding impact.
