Published: 17 September 2026

TL;DR: There are three concrete checks for whether a buy-side agent has negotiated fairly on your behalf. First, retrieve the DealSheet and confirm the agreed CPM is within your mandate parameters. Second, review the agent's decision log to confirm it applied the mandate correctly to the proposals it evaluated. Third, compare the agreed CPMs across a sample of bilateral deals against equivalent open exchange CPMs for the same inventory. Together, these three checks give a buyer an evidence-based answer to the governance question, not a trust-based one.

Deploying a buy-side agent to negotiate advertising deals is a delegation of authority. The agent negotiates deals that the buying organisation is bound by. The natural question that follows: how do you know the agent negotiated well, and within the authority you gave it?

The answer is not to trust that the agent did the right thing. The answer is to have the means to verify it. An agentic marketplace that provides genuine governance gives buyers three verification tools. Each addresses a different dimension of the question.

Check One: The Deal Record

After a campaign, retrieve the DealSheet for each deal the agent agreed. The DealSheet records the agreed CPM, the content category, the inventory parameters, the buyer, the publisher, and the timestamp of agreement.

Check each DealSheet against the relevant mandate parameters:

- Was the agreed CPM at or below the mandate ceiling for the relevant content category? - Did the deal cover an inventory category the mandate permitted? - Was the deal size within the threshold the mandate allowed the agent to agree without escalation?

If the answer to all three questions is yes for every deal, the agent negotiated within its authority. If any deal falls outside the mandate parameters, the marketplace's mandate enforcement did not work as expected, and that needs investigation.

This is the most fundamental check. It answers the question of whether the agent stayed within the authority it was given.

Check Two: The Decision Log

The deal record tells you what the agent agreed. The decision log tells you how the agent reached those decisions: what proposals it received, what parameters it applied to evaluate them, what it declined and why, and what it accepted and why.

Review a sample of the decision log entries:

- Were declined proposals declined because they fell outside mandate parameters? A proposal declined because the CPM was above the ceiling is correct behaviour. A proposal declined for a reason that is not in the mandate is a sign that the agent is applying logic outside its governance framework. - Were accepted proposals accepted because they met mandate parameters? An acceptance should map to a clear mandate parameter: this CPM is within the ceiling, this content category is on the approved list. - Were any proposals escalated to human review? Check that the escalation triggers match the approval thresholds in the mandate.

The decision log check answers a different question from the deal record check. The deal record confirms the outcomes were within mandate. The decision log confirms the decision process was governed by the mandate, not just that the outcomes happened to be acceptable.

Check Three: Market Comparison

A mandate that is well-calibrated to market conditions should produce CPMs that are competitive with equivalent inventory bought through other channels. This is not a governance check in the strict sense. It is a commercial performance check.

Take a sample of bilateral deals from the campaign. For each deal, find the equivalent open exchange CPM for the same or comparable inventory during the same period. Compare the bilateral CPMs against the open exchange CPMs.

What to expect: bilateral deals with specific publishers for premium CTV inventory will typically carry CPMs above the open exchange equivalent for the same inventory type, because bilateral deals offer confirmed placement, content adjacency guarantees, and audience certainty that open exchange does not. If your bilateral CPMs are significantly below open exchange for comparable inventory, the likely explanation is that the mandate ceiling was set conservatively and the agent had room to agree at higher CPMs than it needed to.

This check does not tell you the agent did anything wrong. It tells you whether the mandate was calibrated to market conditions. If the deals were within mandate but below market benchmarks, the issue is the mandate, not the agent.

Putting the Three Checks Together

The three checks address three distinct questions. The deal record check answers: did the agent stay within its authority? The decision log check answers: did the agent apply its authority correctly? The market comparison answers: was the mandate calibrated to produce competitive outcomes?

A buyer who can answer all three questions has done genuine governance, not performative governance. The answers do not require trusting the agent or trusting the marketplace. They require retrieving and checking records that the governance structure makes available.

If any check produces an unexpected result, the action is clear. A deal outside mandate parameters: investigate how the mandate enforcement failed. A decision log entry that applies logic not in the mandate: review the agent's decision framework. CPMs significantly above market benchmarks: review whether the mandate ceiling needs adjustment.

What This Requires From the Marketplace

These three checks are only possible if the marketplace provides the underlying access. A buyer who cannot retrieve DealSheets independently cannot do the deal record check. A buyer who cannot access the decision log cannot do the decision process check.

When evaluating an agent-to-agent marketplace, ask explicitly: can I retrieve DealSheets for all deals my agent agreed, independent of the counterparty? Can I access the agent's decision log? If the answers are no, the governance framework is not complete, and the three checks described here are not available to you.