Reffkin’s Riddle

As I write this, Compass CEO, Robert Reffkin. is sitting in a federal courtroom in Chicago. Zillow’s preliminary injunction hearing against MRED and Compass opened this week, and the CEO of the largest brokerage in America is on the witness list. Post-hearing briefs are due July 9th. However Judge Tarp rules, one thing is already clear: this is not where a winning strategy ends up.

So let’s play a game. What would I do if I were Robert Reffkin?

First, some history.

Compass came out of the gate as a technology company that happened to do brokerage. That was the whole pitch. Their former CTO Joseph Sirosh (a Microsoft AI guy, no less) described Compass as a marketplace platform, like Amazon. They IPO’d in April 2021 at $18 a share.

Strategy Number One : We are a tech company

Wall Street looked under the hood and saw a real estate brokerage with a big engineering budget. The stock sank as low as $1.96. It sits around $12 today. Over $900 million went into that platform, and in August 2022 Sirosh was let go as part of a cost reduction program. Strategy number one, dead. And too be fair many brokerages tried the “we are technology company” angle.  Remember Gary Keller of Keller Williams on stage?  Good times.

Strategy Number Two: Inventory differentiation

If you can’t differentiate on tech, differentiate on inventory. In late 2024 Compass rolled out its 3-Phase Marketing Strategy. Start every listing as a Private Exclusive, stair-step to Coming Soon, then (maybe) the MLS. The vision, as I understand it, is billboards that say “Compass has listings Zillow doesn’t.” Make the buyers come to you.

The problem is that the entire plumbing of American real estate is built to work the other way. So Compass went to war with it. The Clear Cooperation fight. Reffkin personally emailed MLS executives asking them to cut Zillow’s feeds. The NWMLS litigation in Washington, where the MLS is now counterclaiming that the 3-Phase program is consumer deception, with a state law on its side. The MRED partnership I covered in “Dracarys,” with Compass possibly subsidizing 100,000 agent memberships. MRED cut Zillow’s feed over nine listings, the TRO 48 hours later, and now this week’s hearing. I wrote about all of it. It’s a lot.  And I hear MRED membership count hasn’t moved a bit. Rob, get ready to buy me that steak dinner.

And while Compass was fighting that war, the biggest distribution player on earth quietly picked a side. I wrote last week about Google building its national home search on MLS data. Every private network, every coming-soon feed, every velvet rope in the business was available to them, and the company that knows more about how people search than anyone alive bet on the open MLS. So here’s the question that should haunt the whole strategy: what is a Private Exclusive worth when Google can’t see it? Strategy number two, dead.

Strategy Number Three: Scale. 

If you control more listings than anyone, maybe the industry has to come to the table. So Compass bought Anywhere. Announced September 2025, closed January 9th, $1.6 billion in stock, and suddenly Reffkin runs Coldwell Banker, Century 21, Sotheby’s, Corcoran, and ERA…awesome…wow. Per the 2026 RealTrends data, Compass did $262 billion in volume and Anywhere did $193 billion. Combined, that’s 47% of everything the top ten brokerages in America sold. The deal sailed through the federal HSR waiting period, but Letitia James’s office opened an antitrust inquiry in June, and there’s now a class action in Florida over Compass’s $475 transaction fee. Hard to argue “consumer choice” in one courtroom while consumers are suing you in another. Scale bought him more courtrooms.  Strategy number three, is on life support.

And here’s the part that should keep him up at night. While Compass was fighting the listing wars, AI ate the moat. Everything changed after Claude Code, and then OpenClaw in January. Any software built before AI isn’t an asset anymore. It’s an anchor. That $900 million platform Wall Street wouldn’t pay a multiple for? An agent with a laptop can now spin up most of it in a weekend. The “we’re a tech company” argument didn’t just fail to convince Wall Street. It expired. Time to pull the plug Robert.

Which brings us to the fever dream. Wall Street has always had one about real estate: why are we paying agents so much? It reminds me of Travis Kalanick saying the quiet part out loud, that Uber would be a better business without drivers. Now with AI, the fever dream has a new pitch deck. Fewer humans for everything, including real estate.

But there’s a graveyard full of proptech companies that bet against the agent. And they all died for the same reason. A home is the biggest purchase most people ever make. It’s emotional. It’s infrequent, so nobody gets good at it. You live there. You raise your kids there. People want a trusted human for that, and I don’t think AI changes it. I think AI makes it more true, because when a machine can do everything else, the trusted human is the scarce thing.

So here’s what I’d do if I were Robert Reffkin. Stop differentiating the listings. Differentiate the agents.

And I mean with teeth, because “we have the best agents” is what every brokerage has claimed since the dawn of the yard sign. Here’s the version with teeth: radical transparency. Publish every Compass agent’s track record. Actual days on market. Sale-to-list ratio. How often they double-end. Nick Aufenkamp’s whole critique of the private listings movement is that it hides the negative insights buyers need. Imagine Compass answering that critique with more information instead of less. “We hide nothing, including exactly how good our agents are” is a strategy. It also happens to be the perfect reversal for a company currently in federal court defending its right to hide days on market.

And if you want proof that nobody else in this industry will follow him there, NAR just provided it.

Two weeks ago at NAR’s midyear meetings, the board killed a proposal requiring REALTORS® to disclose to clients when they lack knowledge about a property type or geographic area. The Professional Standards Committee spent two years crafting it. The commitee chair called it something “we all should be doing anyway.” The board debated it for an hour and sent it back to committee to die. Here’s the line from a Florida director that tells you everything:

“We do not need to create ‘ahas’ or ‘gotchas’ in the Code of Ethics. The moment that we disclose [that lack of knowledge] you’ve lost that trust with the consumer for the rest of your relationship with them.”

Read that again. The National Association of REALTORS® believes that if its members are honest with consumers, consumers will stop trusting them. They said the quiet part out loud, in the official minutes. And remember the debate on disclosure of referral fees?  Oy vey.

For as long as I’ve been in this business, people have talked about “raising the bar.” Make it harder to become a REALTOR®. Raise the standards, raise the professionalism, and the public’s trust follows. Great speech. It gets a standing ovation at every conference and dies in every committee, and the knowledge-gaps vote is just the latest body. Here’s why it always dies: NAR is paid by the head. At $156 a year times 1.44 million members, every agent who can’t clear a higher bar is revenue walking out the door. To NAR, a bad agent and a great agent are worth exactly the same $156. The organization that profits from more agents can never be the organization that demands better ones.

But you know who doesn’t get paid by the head? Most real estate brokerages. Compass makes money when transactions close, and in an AI world it makes more money when fewer, better agents close more of them. Reffkin is the one guy in the industry whose economics actually allow him to raise the bar. Thirty years of conference talk, and the manifestation of “raising the bar” turns out to be a brokerage, because it was never going to be the trade association selling memberships at the door.

Which brings me to the billboard I’d actually buy if I were Reffkin: “Compass agents are NOT REALTORS®.”

Think about what the REALTOR® brand carries right now. A $418 million Sitzer/Burnett settlement after a jury found the whole commission structure was rigged. A president who resigned days after a New York Times story detailing years of sexual harassment allegations. And now a trade association that formally voted against telling clients the truth about what its members don’t know. NAR’s own budget tells you where this is heading: membership sits at 1.44 million and they’re budgeting for 1.2 million, which means NAR is planning for a quarter of a million people to walk. Reffkin wouldn’t be starting an exodus. He’d be getting in front of one that NAR’s own finance committee already sees coming.

Yes, “Realtor” is synonymous with real estate agent. It’s Kleenex.  That’s exactly why breaking from it is a story. Nobody writes about a brokerage quitting a trade association. Everybody writes about the largest brokerage in America announcing its agents are held to a higher standard than the one actual REALTORS® just voted to reject. And the exit ramp already exists. The ARA just landed RE/MAX the same week NAR subpoenaed them. An alternative association with franchise-scale membership and a chip on its shoulder is sitting right there.

(The honest hedge: yeah, I know the mechanics are messy. Anywhere’s brands are franchises, and Reffkin doesn’t own the franchisees or their NAR memberships. This starts as a Compass-brand move or it doesn’t start at all.)

There’s one more asset hiding in plain sight, and it’s the one I keep writing about. Everybody in this industry is fighting over supply. The next fight is demand. I wrote Monday about Gitcha publishing structured buyer demand into the MLS, and 48 hours later the M&A started. Who knows more buyers than the company with 330,000 agents? Compass is sitting on the largest book of buyer demand in America and treating it like a Rolodex.

Now, the riddle. Reffkin just spent $1.6 billion to own the most agents. Raising the bar implies standards, accountability, culling. You don’t get to be the biggest and the best at the same time, and in an AI world where consumers will only pay for agents worth paying for, best beats biggest. That’s an uncomfortable math problem for a guy who just bought 330,000 of them.

Step back and every Compass strategy has been the same strategy: an attempt to be something other than a real estate brokerage. Tech company. Inventory gatekeeper. Industry landlord. Wall Street told him what Compass was back in 2021.

The winning move is to finally believe them.