UNDER ALL IS THE DATA | SPECIAL EDITION #002
Austin Runs the Experiment
We argued a revenue test for MLS data access cannot survive. Unlock MLS just enacted one. Now the industry gets an answer instead of an argument.
By Saul Klein and John Reilly August 2026
Writers who make predictions usually get years before anyone checks. We got about a week.
In this series we argued that the future of MLS governance belongs to the license, that the lawful way to separate practicing brokers from data harvesting platforms is to measure the use rather than the user, and that the tempting alternative, a revenue composition test, fails three independent ways. We then watched CRMLS circulate a draft that differentiates mostly by use. And on August 20, Unlock MLS in Austin announced the other branch of the experiment: an adopted rule, effective September 15, 2026, that assigns data feed access partly by revenue. The design space we mapped as theory is now bracketed by two live instruments, one drafted, one enacted. The industry is about to get an answer instead of an argument.
What Austin Did
Unlock MLS rewrote its definition of a participant into three categories. A Base Service Brokerage uses the MLS’s core tools but receives no feed of other brokerages’ listings; Unlock says most of its subscribers sit here, which is itself a fact worth pausing on, because it means most brokerages never touch the payload this whole debate is about. A Direct Services Brokerage receives a feed, must earn the majority of its revenue from directly representing buyers and sellers in transactions rather than merely acting as a referral source, and may use the feed only for recognized purposes such as comparative market analyses and back office systems. A Platform Brokerage receives a feed but fails the Direct Services conditions, either because of where its revenue comes from or because of what it does with the data, and beginning September 15 it must execute a commercial license and pay a fee.
Unlock’s framing of that fee deserves to be quoted precisely, because it will be. The fee covers the cost of administering feed access, and it will “act as a stand-in for the data contribution this type of brokerage isn’t making.” Hold both halves of that sentence in mind. They pull in opposite directions, and the gap between them is where this rule will be tested.
The Two Branches of the Experiment
Put Austin next to the CRMLS draft and you can see the whole design space. Both proceed from the same diagnosis this series has pressed: the cooperative assumed every participant was both a producer and a consumer of listing data, the assumption broke, and the corrective instrument is the license. Both build a two track structure, participant access on one side, commercial licensing on the other. They diverge on the trigger.
CRMLS assigns the track by activity and use: direct fiduciary representation qualifies, referral only activity does not, and every restriction in the draft operates on what is done with the data. Unlock assigns the track by a hybrid trigger, and half of that hybrid is the thing we said could not be done: a revenue test. Majority of revenue from direct representation is the “half your revenue from closed transactions” proposal, cleaned up and adopted. The other half of the trigger, use beyond recognized purposes, is the lawful differentiator this series has advocated from the start. Austin built a rule with one foot on each side of the line.
Why Austin’s Version Is Harder to Kill Than the Naive One
Nobody is denied. A Platform Brokerage still gets the feed. It is rerouted to a paid commercial track, not excluded. Unlock will say this is not an eligibility screen at all, merely track assignment inside a licensing structure, which is precisely the architecture we have argued survives scrutiny. The counterargument writes itself: a surcharge triggered by revenue composition is the same discrimination whether you style it as exclusion or as pricing, and the Exhibit B Note’s bar on level of service tests does not evaporate because the penalty is a fee instead of a denial. But the reroute structure genuinely changes the litigation posture. A plaintiff who still has the data argues about price, not about access, and price cases are harder.
The hybrid trigger gives Unlock a retreat. If challenged, expect Unlock to lean entirely on the use prong and describe the revenue prong as an administrable proxy for it. Whether a proxy that operates on identity can borrow the legitimacy of a test that operates on use is, we think, the precise question a court would have to answer, and it has not been answered before.
The NAR tether is already loose. Unlock decoupled MLS access from REALTOR® membership in 2025, with NAR publicly deferring to MLS discretion on participation requirements. An MLS that has already absorbed that friction is signaling it will absorb Handbook conformity friction too. Which relocates the binding constraint from NAR policy to antitrust law directly. And here geography matters: Austin sits in the Fifth Circuit, where Realty Multi-List, the case requiring MLS access criteria to be reasonably necessary and narrowly tailored, is not persuasive authority but controlling precedent. The one court of appeals whose law an MLS revenue screen must survive is the one Austin lives under.
The Sentence That Will Be Quoted Back
Which brings us back to the fee rationale. “Cost of administering access” is the defensible framing; cost based, nondiscriminatory pricing is what the decree era framework requires and what survives. “A stand-in for the data contribution this type of brokerage isn’t making” is contribution based pricing stated out loud, the exact instrument the cost principle exists to prohibit, offered as the official justification, in the same breath as its own defense. Every MLS executive drafting in this space should learn the lesson at Austin’s expense: the rule you write matters, and the rationale you publish matters just as much. Rationale documents are discovery. If your fee is cost based, say only that it is cost based, and make it true.
What We Will Learn, and When
Part 7 of this series is now falsifiable on a schedule. Watch four things. Whether the revenue prong draws a challenge or enforcement interest within a year, which would be the clearest signal available on revenue based track assignment. Whether the challenge, if it comes, attacks the prong or the published rationale, our money is on the rationale. Whether other MLSs copy Austin’s tiers or CRMLS’s use rules first; the market’s choice between them will tell you whether the industry can distinguish governing use from governing people. And how the fine print treats the populations the naive revenue test embarrasses itself on, appraisers first among them, because a strict revenue from representation screen sweeps every appraiser participant in the market, and how Austin handled that detail will tell you how carefully the rest was drafted.
Read the adopted redline before quoting it; announcements compress, and the operative language is what counts. But the shape of the moment is already clear. Two of the most consequential MLSs in the country looked at the broken symmetry this series described, the firms that will not put inventory in and the entities that only take inventory out, and both reached for the license. One reached for use. One reached for revenue, with a use prong as a backstop and a reroute instead of a door. The next twelve months will tell the industry which instrument the law will let it keep.
Under all of it, as always, is the data. Austin just put a price on not contributing any. Whether that price is a license fee or a legal theory is the experiment, and it is running now.
The views expressed here are the authors’ own and do not represent the positions of any organization either of them serves. This analysis is based on Unlock MLS’s August 20, 2026 announcement and contemporaneous trade reporting, and on the circulated July 2026 draft of CRMLS Rule 19; the adopted Unlock redline should be consulted for operative language, and drafts may change. This article is commentary, not legal advice.
