Two brokers, one house, one commission. The first showed the buyer the property in March and kept in touch through a cold spring; the second wrote the offer in June, after the buyer called a number on a yard sign. Both believe they earned the fee. Real estate has had this argument so often that it built a court for it, and the court's first rule is that there is no rule. In 1973 the National Association of Realtors' board adopted what became Official Interpretation 31, which forbids any local rule that would "predetermine entitlement" to a commission. Every dispute is decided on its facts, at a hearing.
What the panel looks for has a name: procuring cause, which the guidelines define as "the uninterrupted series of causal events which results in the successful transaction." The questions follow from it. Who first introduced the buyer, and when, and how? Did that introduction start a chain that ran unbroken to the closing, or was the chain interrupted, and by whom? Did the first broker keep in contact, or could the buyer reasonably have concluded he had walked away? The burden of proof rests on the party who asks for the money, and the standard is the preponderance of the evidence. Splitting the fee is allowed, but the guidelines call it the exception, for the rare sale that "would have resulted only through the combined efforts of both parties."
Notice what the panel never asks: who was standing nearest the door on closing day.
The Broker Nobody Can Subpoena
Marketing has spent twenty years answering the question real estate refused to answer. Last-click attribution is a predetermined rule of exactly the kind Interpretation 31 bans: it pays whoever held the buyer's hand at the end and calls the matter settled. It survived because the chain of events was mostly visible. An ad was served, a link was clicked, a cookie was set, a form was filled in. The rule was crude, but the evidence was in the room.
AI assistants have taken a witness out of the room. On Thursday the Interactive Advertising Bureau published its AI-Era Attribution Blueprint, and the reason it gives is plain: as AI takes on more of the purchase journey, "the clicks and referrals that attribution depends on are thinning out or disappearing." According to the IAB's own outlook research, 45% of ad buyers name comparing AI-driven and traditional journeys as their top measurement challenge, and 44% rely on branded search and direct traffic as proxies for AI influence they cannot track. Similarweb data cited in the IAB's report, as Digiday reported, found that people who received a brand recommendation from an AI assistant were 2.5 times as likely to visit that brand's website within seven days, yet only 8.8% of those visits came directly from an AI referral. Most of the rest arrived through search, and a fifth arrived direct.
That is the first broker, invisible. The buyer heard the name in a conversation the brand will never read, then typed it into a search box, and the search box collected the commission. The IAB's Caroline Giegerich named the cost precisely: AI-influenced sales reported as search or direct traffic "can send the next marketing investment to the wrong channel." Budget follows credit, and credit is following the yard sign.
A Hearing, Not a Formula
What makes the blueprint worth a leader's attention is that it does not pretend to be a better formula. It reads like an arbitration manual. It asks three questions of every conversion: who or what created the intent, who or what made the decision, and who or what completed the purchase. The answers sort each sale into one of three pathways, Human-Owned, AI-Assisted or AI-Executed, and the pathway decides who may make a decision claim, who may make a contributory claim, and what evidence each claim requires. Then comes the sentence that could have been lifted from the Realtors' guidelines: "A claim is not credit; it is the right to be evaluated for credit."
The trade press has been fair about its limits. MarTech's account notes that much of the evidence needed to establish AI's role is either never collected or stays inside the AI platforms, and that the framework creates a common vocabulary without solving the shortage of observable data. That is true, and it is not a reason to wait. Arbitration panels have never had perfect evidence either. They have had standards for weighing what evidence exists, and a rule about who carries the burden.
Run Your Own Hearing
A marketing leader can adopt those standards this quarter, before any platform volunteers a single new field. The first is to stop treating any channel's self-report as a verdict. The blueprint asks agencies to separate platform-reported claims from independently verified ones, and a brand should hold its own team and its own vendors to the same line: a platform that grades its own contribution is a broker testifying in his own case. The second is to put the burden of proof where the money is. A channel that wants a larger share of next year's budget should be able to show the chain, not merely its position at the end of it; a claim that rests only on evidence the claimant will not disclose has not met the standard.
The third is to estimate, honestly and roughly, how much of your business already runs through each pathway, which the blueprint asks brands to do. That estimate will not come from a dashboard. It comes from asking buyers how they found you, in their own words, on the form and on the first sales call; from reading branded search and direct traffic as a symptom of introductions made elsewhere rather than as demand your search team created; and from watching whether the names that arrive by those routes match the questions people are putting to assistants about your category. None of it is precise. All of it is closer to the truth than a rule that pays the yard sign every time.
The fourth is the hardest, because it concerns the budget rather than the report. If an assistant is introducing buyers to you, then the work that earns those introductions, the clarity of what you publish, the reputation others publish about you, the specifics an assistant can repeat, is doing the job of the first broker. It deserves funding on the evidence of the chain it starts, even when it will never be standing at the door.
Real estate learned long ago that the person who opens the door is rarely the person who brought the buyer to the street. The commission belongs to whoever set the sale in motion and kept it moving. Marketing is about to relearn that lesson with a witness it cannot call, and the teams that do well will be the ones that hold a hearing instead of trusting a rule.
