A neutral ad marketplace is one that does not have a financial stake in which side of a deal wins or in how much value the deal leaks on the way through. Most marketplaces are not neutral in that sense, and "neutral" has become a word vendors apply to themselves without earning. This piece defines neutrality in terms a buyer can actually check, technically, contractually, and operationally, and gives a checklist for testing whether a marketplace calling itself neutral is telling the truth.
TL;DR. Neutrality is not a tone or a brand value; it is a structural property, and it has three testable forms. Technical neutrality: does the platform give both sides of a deal the same information and the same record, or does it privilege one side's view? Contractual neutrality: does the platform make money in ways that align with the deal working, or does it profit from opacity and from favouring one party? Operational neutrality: when buyer and seller disagree, does the platform have a stake in the outcome, and can it be audited by both? A marketplace that passes all three is neutral in a way that survives scrutiny. One that passes none is using the word as marketing.
Why does neutrality matter, and why is the word suspect?
Neutrality matters because a marketplace sits between two parties whose interests do not fully align, and the marketplace's own incentives determine whether it serves both or quietly favours one. If the marketplace profits from fees that are easier to hide from one side, or from routing that benefits the party paying it more, then its neutrality is a claim contradicted by its own business model. The buyer's practical stake is direct: a non-neutral marketplace is one where value can leak toward the marketplace or its preferred counterparty, at the buyer's expense.
The word is suspect because it has become free to use. Any platform can call itself neutral, and many do, because the term sounds like trustworthiness and costs nothing to assert. The programmatic supply chain's long-documented transparency problems, where large shares of ad spend disappear into intermediary fees and opaque routing, are in large part the result of intermediaries whose incentives were not neutral while their language often was. So a buyer should treat "neutral" not as information but as a claim to be tested, and the test is whether the neutrality is structural or merely stated.
Technical neutrality: does both sides see the same thing?
Technical neutrality is about information symmetry. A neutral marketplace gives the buyer's side and the seller's side the same view of the deal, the same data, and, critically, the same record of what was agreed. A non-neutral one privileges one side's view, or keeps the authoritative record in a place one party controls, which means the other party is trusting a counterparty's version of events.
This is where agentic advertising sharpens the question. When agents on both sides transact, each tends to write the deal into its own system, and those records drift apart. Published simulation research into agentic reconciliation found that two agents which had just agreed the same deal recorded its terms differently in the large majority of cases, with both sides passing standard reconciliation checks as the divergence compounded. This is the practical form of a point both AdCP and IAB Tech Lab have made on the record: an agent asserting a transaction occurred is not a receipt, and only a deterministic, auditable record both sides can reference settles what the deal actually was. That research models the specification architecture rather than any live platform, but it shows why a shared record is the technical heart of neutrality: if the two sides do not reference one record, the marketplace cannot be neutral about a deal, because there is no single neutral account of what the deal was. The checkable question is simple: do both sides read from and write to the same record, or does each keep its own?
Contractual neutrality: does the platform profit from the deal working?
Contractual neutrality is about how the marketplace makes money, and whether that aligns with the deal succeeding or with obscuring it. A neutral marketplace charges in ways that are visible to both sides and that do not depend on hiding value from either. A non-neutral one profits from opacity, from fees that are hard to see, from spreads it does not disclose, or from favouring the party that pays it more, so that its revenue rises when the buyer's or seller's understanding of the deal falls.
The checkable questions here are contractual and specific. Are the platform's fees disclosed, in full, to both sides? Does the platform take an undisclosed margin or spread on transactions? Does it have commercial arrangements that would incline it toward one counterparty over another? A marketplace whose published fee structure is the same regardless of which side you ask, and whose revenue does not depend on either side knowing less, has a business model consistent with neutrality. One that cannot or will not disclose how it profits from a given deal is asking to be trusted on exactly the point where trust should be earned.
Operational neutrality: who has a stake when there is a dispute?
Operational neutrality is about what happens when things go wrong. Deals produce disagreements, over delivery, over terms, over what was actually agreed, and a neutral marketplace has no stake in how those disagreements resolve, while a non-neutral one benefits from resolving them in a particular direction. The question is whether the platform is a disinterested venue or an interested party wearing the venue's clothes.
The checkable version asks who adjudicates and who can audit. When buyer and seller disagree, does the marketplace decide the outcome, and if so, does it gain from deciding one way? Can both parties independently audit the record to establish what happened, or must they accept the platform's account? A marketplace that lets both sides verify the facts against a shared record, and that does not profit from the resolution going one way, is operationally neutral. One that controls both the record and the adjudication, and has an interest in the result, is a referee with a bet on the game.
The buyer's neutrality checklist
Run a marketplace claiming neutrality against these questions. Each is answerable, and the pattern of answers tells you whether the neutrality is structural or cosmetic.
Do the buyer's side and the seller's side see the same information about a deal, or does the platform privilege one view?
Is there a single shared record of each deal that both sides reference, or does each party keep its own books?
Are all of the platform's fees disclosed, in full, to both sides of the transaction?
Does the platform take any undisclosed margin, spread, or markup on deals?
Does the platform have commercial arrangements that would incline it toward one counterparty over another?
When buyer and seller disagree, can both independently audit the record, or must they accept the platform's version?
Does the platform gain financially from a dispute resolving in a particular direction?
Is the platform's revenue model consistent with deals being transparent, or does it depend on one side knowing less?
A marketplace that answers all of these in the neutral direction, symmetric information, a shared record, full fee disclosure, no hidden margin, no favoured counterparty, mutual audit, no stake in disputes, is neutral in a way a buyer can defend. A marketplace that answers most of them the other way is not neutral, whatever its marketing says, and the gap between its language and its answers is precisely the risk.
Why neutrality is worth insisting on now
The agentic transition raises the cost of getting neutrality wrong. When humans transacted slowly, a non-neutral marketplace leaked value at human speed, and relationships and manual checks caught some of it. When agents transact continuously, a non-neutral marketplace leaks value continuously, and the agents, acting on whatever information they are given, have no relationship and no instinct to catch it. Agents will optimise against the record they are handed, so if that record is controlled by an interested party, the agents will efficiently serve that party's interest without ever knowing they are doing so.
That is the reason to treat neutrality as a structural requirement rather than a nice-to-have, and to test it with the checklist rather than accept it as a claim. In a market where decisions are increasingly made by software acting on records it did not create, the neutrality of whoever holds the record stops being a matter of good faith and becomes a matter of architecture, which is why the IAB Tech Lab's agentic standards work centres on canonical, auditable transaction states rather than natural-language confirmation. A buyer's protection is not the marketplace's promise to be fair. It is the marketplace's inability, by design, to be otherwise.
This article references the Association of National Advertisers' programmatic transparency research on supply-chain fees and opacity, and cites published simulation research into agentic deal reconciliation conducted by Alkimi, and draws on the IAB Tech Lab's Agentic Advertising Management Protocols (AAMP). The simulation models the current specification architecture and is not an assessment of any specific production platform.