Ad verification technology

Ad verification technology checks whether a digital ad ran under the conditions a buyer expected. Depending on the campaign and its integrations, it can assess viewability, invalid traffic, content suitability and placement. A useful report shows both the finding and how much inventory the system could assess. An impression marked “unknown” has not passed a check.

What the checks answer

These are separate decisions. An ad can be viewable beside content a brand would avoid. An ad can also be non-viewable without being fraudulent. Avoid combining the findings into an unexplained “quality” score.

Ad Verification Checks: What They Measure and Key Limitations

Where verification fits

Some controls act before a bid or ad delivery; others observe and report after an impression. A control applied before delivery may prevent a known unwanted placement, but it cannot report a later on-screen event it never observed. Measurement after delivery depends on the signals available to it.

Specify whether each rule blocks an ad, flags a finding or only reports it.

Technical access matters. IAB Tech Lab’s Open Measurement SDK provides common code through which integrated sites and apps can expose signals to measurement partners. Integration does not mean every publisher, player or device supplies every signal.

For video, the VAST specification supports verification resources, but a particular player must execute them. In connected TV, server-side ad insertion and limited player signals can complicate measurement and invalid-traffic assessment.

Treat coverage as a matrix. For each metric, format, platform and publisher, ask whether the provider observes the relevant signal, uses supplied data or cannot make a determination. Request counts for each state. For content suitability, distinguish inspection of a particular page or content item from a broad site or app label.

What a shared measurement standard can add

IAB Tech Lab’s Open Measurement SDK is designed to make impression measurement more consistent across connected TV, mobile and online video. Without a shared approach, platforms can collect basic signals such as on-screen placement, overlays and video quartiles using different methods, making results harder to compare.

The signals described for the OM SDK include when an ad appears on screen, the percentage in view and how long it remains there. When assessing reporting, check which of these signals are available for the campaign’s inventory, rather than assuming that use of a common standard means every signal is present.

The implementation uses the Open Measurement Interface Definition (OMID): an integrated site or app can send measurement signals to the API, and measurement providers can place tags to collect them. The OM SDK is designed to let companies work with multiple verification partners through one implementation, rather than separate custom integrations for each provider.

Benefits and Risks of Using Shared Measurement Standards (OM SDK)

Choose the report before choosing a provider

Define the decision the report must support. For display and video, request viewable, non-viewable and undetermined counts separately by format.

For invalid traffic, ask which general and sophisticated detection methods apply, when findings can change and how filtered counts reconcile to the campaign total. For adjacency, ask what content the provider can inspect and what it infers from publisher signals.

A useful evaluation request specifies:

Compare providers against the same requirements and inventory where a pilot is possible. Accreditation of one service or metric does not establish coverage of every format or suitability category.

Check the video implementation

For video, the IAB Tech Lab’s VAST specification describes expected behaviour for video players responding to ad requests. VAST 4.0 separated media files from verification and interactive code, while VAST 4.1 added support for ad verification with Open Measurement. The specification’s capabilities still depend on the player and integration used in a campaign.

Include the relevant VAST version and verification execution in the technical questions for a video buy. Ask whether the player can execute the verification resource and whether the reporting provider receives the intended measurement signals; a specification feature alone does not confirm that a particular campaign will expose them.

For connected TV and other video environments, compare providers using the same inventory and ask them to identify which signals they actually receive. This makes it easier to distinguish a measurement limitation in the integration from a campaign finding, and to set expectations for what the report can and cannot establish.

How to Evaluate a Video Ad Verification Implementation

Make the findings usable

Agree who receives alerts, which discrepancies merit investigation and what data the buyer and seller can exchange. Review tracking early so a failed integration does not quietly fill the unknown bucket for an entire campaign. Record rule changes: a stricter suitability policy can raise flagged volume even when the underlying inventory has not changed.

Use verification to support a specific decision: keep a placement, change a control, investigate a source or accept a documented limitation. Its reports describe conditions within their stated coverage. They do not, by themselves, establish attention or campaign effectiveness.