What MCP means for the Real Estate Brokerage you run

If you run a brokerage and you've ever felt stuck inside a software contract you can't get out of, paying for a system you can't get your data out of, watching your operations team spend half their week wrestling with tools that should be helping them, I want you to know two things.

First: it isn't your fault. For most of the last few decades, that was the deal. You signed it because everyone signed it. The other options were worse.

Second: the deal is changing. Quietly, and faster than most of the industry has noticed.

I've been building software for real estate brokerages for over fifteen years. In all that time, I've never seen a shift this clean. I want to walk through what's actually different now, and what I'd be doing about it if I were sitting in your chair.

What changed

There's a new piece of plumbing called MCP (Model Context Protocol). The name is boring on purpose. What it does is not.

In plain language, MCP is a standard way for an AI model like Claude to talk directly to the software you already use. Your commission system. Your CRM. Your transaction management tool. Your accounting platform. Instead of you logging into five tools to answer one question, you ask the question once, and the AI goes and gets the answer from wherever it lives.

That's it. That's the whole idea. But the consequences are big.

It means your brokerage data isn't trapped inside a vendor's product anymore. It means the AI working for you doesn't care which tools you picked. It means the thing that used to require three contractors, a BI tool, and a six-month implementation now takes a conversation.

We rebuilt Loft47 to work this way. Not because it was trendy. Because once you see it working, you can't unsee how broken the old way was.

Legacy players are not going to like this

There are still software vendors in this industry selling brokerages on multi year contracts for products that run on desktop installs or RDP. Actual desktop software sold under contracts dense enough to make switching cost more than staying.

Those companies survived this long for one reason: switching was painful, and they made sure of it. Their pitch was always the same. "Use the all-in-one suite. Everything talks to each other. You won't need anything else."

You did need something else. You always did. And every brokerage I've worked with has the spreadsheet to prove it. The shadow system the office manager built to make the "all-in-one" actually work.

What MCP changes is the floor. The leverage your vendor had, that your data was hard to get to, that nothing else could talk to it, that switching was scary, that leverage was the whole product for some of these companies. When the data is accessible and the integrations are trivial, what's left? In a lot of cases, not enough.

I'm not gleeful about this. Some of those companies have been around longer than I have. But the brokerage owners who stuck with them weren't being loyal. They were being held hostage. 

Your margins do not have room for this anymore

Commission splits are getting compressed from both ends. Agents are negotiating harder. Buyers and sellers are asking sharper questions about value. The brokerages I've watched come through the last two years strongest are the ones who got serious about overhead and not by cutting people who matter, but by getting the operational work that doesn't need a human off humans' desks.

Reconciling trust accounts. Chasing missing documents. Answering an agent's "why is my commission this number" question for the forty seventh time this month. Pulling reports for ownership. Running the same compliance checks before every close.

None of that work needs a person. All of it currently sits on a person in your office, probably one of your most expensive operations hires, because the software you bought a decade ago can't do it without one.

When you put your data behind MCP and let a model like Claude work on it, that work moves. Not all at once, and not magically. But the financial audit that used to take your back office three days starts taking ninety seconds. The deal status questions that used to interrupt your manager twenty times a day start answering themselves. The compliance gap sweep that used to wait until the end of the month happens every morning at 6am, before anyone has logged in.

That's not just a cost-cutting story. That's an operational profitability story. The brokerages who figure this out first are going to look very different from the ones who don't.

Compliance is local. Now your workflow can be, too.

Anyone who's tried to use the same software in Ontario, Alberta, Florida, and California knows what I'm about to say. The rules are not the same. The trust and escrow account requirements are not the same. The disclosure timing isn't the same. The way an excess release works isn't the same. Even the words for the same thing aren't the same.

The "all-in-one" platforms tried to solve this by averaging. Building one workflow that almost works everywhere, which means it really works nowhere. Brokerage operations people have spent the last decade duct-taping around the gaps.

With MCP, the workflow can finally be as specific as your jurisdiction demands. Your Toronto-based brokerage and your Tampa-based brokerage can run on the same data and the same tools, with completely different workflows on top, because the model adapts to whatever you need. The compliance officer for an Alberta brokerage doesn't need to know what's in a Florida disclosure packet. They just need their workflow. That's the version they get.

This is the part of the shift I find most personally exciting. The industry has spent a long time forcing brokerages to fit the software. The software is finally starting to fit the brokerages.

The brokerages that move first are going to be very hard to catch

I don't usually go in for "the future is here" talk. It always sounds like a sales deck. But this is a moment where the gap between the brokerages who adopt and the brokerages who wait is going to widen quickly.

For the simple reason that a brokerage running this way has lower overhead, faster close cycles, cleaner compliance, better retention coaching, and far more useful conversations with agents about their numbers. That advantage compounds. 

What I'd be doing if I were running your brokerage

I'd stop signing multi-year contracts with anyone whose product can't expose your data to the tools you choose. That's the new baseline. If a vendor can't or won't, you're being asked to bet on their product being the last system you'll ever need. Nobody should take that bet.

I'd ask my operations team where the time is going. Most brokerages I work with discover that thirty to forty percent of their back office hours are on work an AI agent could be doing today, on data they already have. That's a huge opportunity.

And I'd start small, maybe pick one painful weekly task to automate. See what changes. Then pick another.

You don't have to rebuild your stack. You just have to stop renting access to your own data.

If you want to see what this looks like running on a real brokerage, that's what we're showing on Loft47 TV. We're building one of these every week, on real-world data, end to end. Come watch. And if any of it lines up with something you're trying to solve, send me a message. I’d rather start with a conversation than a solution.

‍

‍

Sasha Hryciuk is the founder of Loft47. She's been building software for real estate brokerages since before "proptech" was a category.

‍