PMPs now carry 99.2% of programmatic CTV spend. That number was supposed to be the end of the CTV quality debate: move everything into private deals, away from the open market’s chaos, and the premium label takes care of itself. This week brought the argument that it did no such thing — and the evidence to back it.
The case comes from Basis’ Ayse Pamuk, and it is aimed at the industry’s favourite framing: the PMP-versus-open-market divide is a distraction, because the deal type stopped being a quality signal the moment everything became a deal. What actually separates outcomes is quality control inside the deal — and inside many deals carrying premium labels, that control is absent. The mechanism is mundane and structural: deals blend genuinely premium inventory with lower-quality supply to hit volume commitments. The buyer sees one deal ID, one label, one blended price. The premium content inside is carrying passengers.
The numbers put scale on it. Without safeguards, more than 25% of CTV impressions may fail minimum quality standards — a figure from the vendor sphere that should be treated as directional rather than audited, but which matches what anyone who has pulled app-bundle-level delivery reports on a “premium” deal will recognise. The harder evidence is the ANA’s: top-performing advertisers converted 54% of spend into qualified impressions, versus 32.1% for the rest — despite similar PMP usage. Same deal type, same “premium” label, nearly 70% more delivered value per dollar. The difference was not the acronym. It was the controls.
01The blend is a pricing mechanism — and you’re on the wrong side of it
For a publisher holding genuinely premium CTV supply, this story reads like a measurement problem. It is actually a pricing problem. When premium and mid-tier inventory blur together inside one deal ID, the deal clears at a blended price — which by definition is below what the premium portion is worth on its own. The owner of the genuinely premium impressions is subsidising the owner of the filler. Every quarter this persists, the market’s reference price for “premium CTV” drifts toward the blend, and the discount compounds.
The volume commitment is where the dilution originates, and it deserves to be named as the villain. A buyer asks for scale; the deal’s premium supply cannot fill it; and someone in the chain — an SSP, a reseller, sometimes the publisher’s own ad ops under quarterly pressure — quietly widens the definition of what the deal contains. Nobody decided to defraud anyone. The label just stopped meaning anything, one impression at a time. That is how a market where 99.2% of spend is “private” and “premium” still fails a quarter of its impressions on basic quality.
The ANA gap is simultaneously the indictment and the sales opportunity. If top advertisers extract 54 cents of qualified delivery per dollar and the median extracts 32, then a deal that verifiably delivers the top-quartile experience is worth a materially higher CPM — and there are now published, buyer-side numbers to anchor that negotiation. Clean supply has never had better ammunition. It just has to be provably clean.
02Why this matters for publishers
| Blended deals tax your best inventory | If your premium CTV supply shares a deal ID with mid-tier inventory — yours or, worse, someone else's — it is clearing at a blended price. You are funding the discount on supply you may not even own. |
|---|---|
| The premium label has stopped carrying information | At 99.2% PMP penetration, "it's in a PMP" tells a buyer nothing. Verifiable composition — content tier, app bundle transparency, delivery environment — is the new premium signal, and it is scarce. |
| Buyers are being taught to audit | The ANA's 54%-versus-32.1% gap gives every sophisticated buyer a reason to demand impression-level accountability. Publishers who can pass that audit convert it into pricing power; publishers who cannot will discover what their deals actually contain at the worst possible moment. |
| The volume-commitment trap is structural | Any deal whose volume ask exceeds your genuinely premium capacity will be filled with something. If you do not control what, the chain will decide for you. |
CTV spent five years solving the wrong problem: it moved 99.2% of spend into private deals and called that quality, when the quality was only ever as good as the controls inside each deal.
03What publishers should do
04The bottom line
CTV spent five years solving the wrong problem: it moved 99.2% of spend into private deals and called that quality, when the quality was only ever as good as the controls inside each deal. The market that emerges from this correction will price verifiable composition, not deal-type acronyms — and that is unambiguously good news for publishers whose supply is genuinely premium, because they are currently the ones paying for the blend. Separate the tiers, document the composition, and charge for the proof. The buyers who matter have just been handed the numbers that justify paying for it.