Stop fighting the split war. Cut the cost of running the house.
Commission pressure and rich agent splits are squeezing brokerage margins from both sides. The lever that protects your unit economics is not a better split offer. It is a cheaper way to run the brokerage.
Key Takeaways
- The squeeze is real: on June 11, 2026, Inman reported on the quiet financial crisis inside real estate brokerages. Post-settlement commission pressure plus high agent splits are compressing what the house keeps on every deal.
- The trap: the instinct is to compete on splits, which is the one move that makes the squeeze worse. A more generous split wins the recruit and loses the margin you were trying to protect.
- The real lever: AI lowers the per-transaction cost of running the brokerage, the back office, transaction coordination, compliance review, and recruiting admin, so the house can keep a profit at a competitive split instead of beating rivals to zero.
- The takeaway: AI here is not an agent gimmick. It is the thing that fixes unit economics. The broker-owner who cuts the cost to close a file can offer a fair split and still make money, which is a structurally stronger position than the one next door.
The Leveraged Years Briefing. Permalink
The crisis Inman named
On June 11, 2026, Inman published a piece on what it called the quiet financial crisis happening inside real estate brokerages. The shape of it is simple and painful. After the commission settlement changes, the money coming in per transaction is under pressure, while the splits paid out to agents stay high because that is how you keep them. Revenue per deal soft on one side, payout per deal stiff on the other, and the house caught in the middle.
What makes it quiet is that it does not show up as a single bad quarter. It shows up as a slow erosion of the margin on every closed file, the kind that a busy broker-owner can miss while transaction count looks fine. The brokerage can be busier than ever and quietly making less on each deal than it did two years ago.
This is a business-economics problem, not an agent-production problem. The agents may be performing well. The unit economics of the house are the thing under stress.
Why competing on splits is the trap
When margin gets tight, the loudest move in the market is to fight harder on splits. A rival down the street offers ninety, so you offer ninety-five. You win the recruit. You also just signed up to make even less on everything that agent closes.
The split war is a race in the wrong direction. It treats the agent payout as the only variable you can move, when payout is actually the one variable that directly destroys the margin you are trying to save. Every point you give on split comes straight out of the house's economics, and the competitor matches you next week anyway.
The brokerages that survive this do not win the split war. They change the math underneath it. They make the cost of running the house low enough that a competitive split still leaves a profit. That is a defensible position. A higher split offer is not.
Where the cost actually hides
The cost of running a brokerage is not mostly in the split. It is in the work that surrounds every transaction and never shows up on a commission statement. This is where AI earns its place, and it is the part most owners underprice.
- Back office. The coordination, the document handling, the email threads that move a deal from contract to close. Each one is small. Multiply it across every file and it is a real headcount.
- Transaction coordination. Chasing signatures, tracking deadlines, assembling files, flagging what is missing. Highly repetitive, highly rules-based, and a natural fit for AI-assisted workflows that draft, summarize, and remind.
- Compliance review. Reading files for the things that must be there and the things that must not. A model that surfaces gaps and inconsistencies for a human to confirm turns a slow manual pass into a fast review.
- Recruiting and onboarding admin. The repetitive paperwork, scheduling, and follow-up that surrounds bringing on agents, separate from the human judgment of who to recruit.
None of these is glamorous. All of them are cost. And all of them are the kind of structured, repetitive work where AI cuts time without touching the parts that need a licensed human.
How AI fixes the unit economics
The move is to attack the per-transaction cost so the house keeps a profit at a split you can actually offer. A few honest principles.
- Aim AI at the back office before the front office. The agent-facing tools your agents use to write listings and market themselves are useful, and the hub, How real estate runs on AI, covers that ground. But the margin problem is solved in operations, not in marketing copy.
- Use AI to compress, not to replace. The goal is to let a smaller operations team handle more files without a drop in accuracy. A coordinator who closes thirty files a month with AI support instead of eighteen changes your cost per deal without changing your culture.
- Keep a human on every judgment call. Compliance and money decisions stay with a person. AI drafts, summarizes, flags, and reminds. The licensed human confirms. That line is what keeps the savings from turning into liability.
- Measure cost per closed transaction, not just count. The number that tells you whether this is working is what it costs the house to take one file from contract to close. Drive that down and the squeeze loosens at a competitive split.
Your agents will be doing their own AI work on the listing and marketing side. The briefing How real estate agents use Claude AI is the right thing to hand them. Your job as the owner is the layer underneath: the cost of the machine they run on.
The skill under the squeeze
Commission models will keep shifting, and the next settlement headline will land like the last one. The broker-owners who come through this are not the ones who guessed the market right. They are the ones who built a brokerage that makes money at a fair split because the cost of running it is low.
That is an operating skill, and AI is the strongest lever for it in a generation. The Leveraged Real Estate Series teaches the method for putting AI to work on the economics of a brokerage, not just the marketing, and the two minute course quiz will point you to the right starting place for your role.
Frequently Asked Questions
What is causing the brokerage margin squeeze?
According to Inman's June 11, 2026 report, it is the combination of post-settlement commission pressure on revenue per transaction and high agent splits on the payout side. The house gets compressed from both directions, often slowly enough that it is easy to miss while transaction volume looks healthy.
Why not just offer a higher split to keep agents?
Because the split is the variable that directly destroys the margin you are trying to protect. A higher split wins the recruit and shrinks what the house keeps on every deal that agent closes, and competitors match the offer anyway. It worsens the squeeze rather than solving it.
Where does AI actually save a brokerage money?
In the per-transaction cost of running the house: back office, transaction coordination, compliance review, and recruiting and onboarding admin. These are repetitive, rules-based tasks where AI compresses the time and headcount needed, while licensed humans keep every compliance and money decision.
Is this briefing financial or business advice?
No. The Leveraged Years is an education company, not an accounting or business advisory firm. This is a plain language explainer of a current trend in brokerage economics. Treat it as background, and confirm anything that affects your firm's finances or compliance with a qualified professional.