Programmatic HCP campaigns routinely lose a meaningful share of working media to broken identity matches, stale target lists, and impressions that cannot be tied to a verified prescriber. The number rarely appears on a dashboard because the party reporting the dashboard is the party being measured. It belongs in the budget review.
Every HCP media plan rests on one assumption: that the impressions bought against a target list were delivered to the people on it. In practice that assumption depends on an identity chain with several weak links. An NPI has to be resolved to a person, the person to a set of devices and browsers, the devices to an ad opportunity, and the opportunity to a verified delivery. Each link has a failure rate. The failures compound.
Industry practitioners now quote a range for that compounded loss that would be unacceptable in any other line of the budget. Yet most brand teams have never seen the number for their own plan, because it is not a standard field in any vendor report.
Three failure modes account for most of the waste. The first is resolution: the identity graph maps an NPI to a device that belongs to someone else in the household, the practice, or the hospital network. The second is decay: a list built from prescribing data with a lag of months targets clinicians who have moved, retired, or changed specialty, and misses the ones who now see brand-eligible patients. The third is authentication: an environment reports a match against a self-declared profile rather than a verified credential, and the reported rate is the share of impressions the vendor was willing to label, not the share it could prove.
None of these show up as a problem in reach and frequency reporting. The impressions were delivered. The report is accurate. It is simply measuring the wrong thing.
A platform that sells impressions against an HCP list is paid on the volume it can label as matched. Its match methodology, its refresh cadence, and its definition of "verified" are all commercial choices, and each one tilts toward a larger addressable count. That is not misconduct. It is the incentive that the business model creates, and it is exactly why the audit has to sit outside the buy.
The same logic applies to the agentic planning tools now being sold by the same platforms. A system that optimizes toward its own inventory will find its own inventory efficient.
The fix is procedural rather than technical. First, require every vendor to report match rate by method and by channel, with the share of impressions that could not be verified stated separately rather than folded into the total. Second, run an independent match-quality audit quarterly, and rebuild the identity approach when the failure rate crosses the threshold the team has set in advance. Third, replace a static list refreshed a few times a year with a dynamic policy that predicts which clinicians will see eligible patients next quarter, not which ones prescribed last year. Fourth, move first-party matching into a clean room under the brand's own contract, so the match is measured against the brand's data rather than the vendor's claim.
Then put the resulting number into the budget review, as a line, next to the media cost it applies to. Once leadership sees the share of working media that never reached a verified prescriber, the reallocation conversation has a foundation. It becomes a question of what the recovered spend should buy, rather than whether the problem exists.
An independent audit ends with three things: the waste stated as a range in dollars with its assumptions visible, a scorecard that compares every active vendor on the same criteria, and a remediation plan the team can execute in a quarter. Where the client permits, sample checks separate reported quality from observed quality. Where something cannot be known, it stays labeled unknown, and it usually becomes the first remediation item.
The audit is not an argument against programmatic HCP media. It is the precondition for spending on it with confidence.
A fixed-fee audit states the waste in dollars, scores every vendor, and ends in a remediation plan. Twenty minutes to scope it.