How to Audit a Pharma CTV Supply Chain

Trace each impression path

IAB Tech Lab's app-ads.txt specification extends authorized seller declarations to apps, while sellers.json helps identify direct sellers and intermediaries. These transparency tools do not certify content quality or eliminate fraud (ads.txt/app-ads.txt; sellers.json).

Find duplicate and indirect paths

The same app inventory can arrive through several exchanges or resellers. Compare win rate, total cost, quality, and unique reach by path. Prefer a shorter route only when it preserves access and performance; fewer hops are not automatically better if a path adds legitimate value.

Validate the viewing environment

Confirm whether the impression was actually television-like, whether autoplay and sound behavior match expectations, and whether completion is meaningful. Investigate unknown apps, implausible volume, extreme completion patterns, repeated households, and delivery outside approved geography.

Connect quality to outcomes

Supply metrics diagnose whether an impression opportunity was credible. They do not prove business impact. Compare qualified reach and outcome evidence by supply tier, while accounting for audience and cost differences.

Hypothetical finding

A PMP supplies 40% of impressions but 65% of unique households, while a reseller-heavy path generates high completion and repeated frequency. The action may be to cap or remove that path, then monitor whether marginal reach improves. Percentages are illustrative.

Red flags that justify a deeper look

Some patterns in CTV delivery data should trigger an immediate investigation rather than waiting for the monthly audit:

• Unknown or unnamed apps receiving meaningful spend.
• Completion rates near 100% across a path, which can indicate non-human or forced viewing.
• Delivery at unusual hours that does not match normal viewing patterns.
• The same household IDs appearing at very high frequency from one path.
• Sellers missing from app-ads.txt or listed with a different relationship type.
• Sudden volume spikes from a new source after a bid or deal change.

None of these proves fraud on its own. Each is a reason to pull log-level data, check authorization, and ask the seller for an explanation. If the explanation is not satisfactory, pause the path while you investigate. Use the partner escalation framework for anything that looks material, and see how to use ads.txt, sellers.json, and SupplyChain objects together for the tracing steps.

Practical takeaway

CTV audits should end with an allow, monitor, or block decision at the most actionable supply level. The next step is a monthly path table joining seller transparency, placement, fees, household frequency, and outcome evidence.

Editorial note. Analysis and frameworks are the author's own and do not represent Acxiom or any current or former employer, client, or named platform. Examples labeled hypothetical or illustrative are not results from real campaigns. Nothing here is legal, regulatory, or medical advice.