If the agents can all leave on 30 days notice, what is brokerage equity actually worth?

The property management comparison is doing a lot of work in your post and I think it's the right comparison, which is why the answer is mostly the earnout. If the thing you're buying can dissolve, the seller has to stay exposed to whether it dissolves. Anything else is the seller selling you their own retention risk at a premium.

I've been on the wrong side of a version of this in a different business. The earnout sounds clean until you argue about the definition for two years. Retained production measured how? If a producer stays but their volume drops 40% because the market slowed, does the seller eat that? If they leave and the buyer's own recruit replaces the revenue, is the seller paid?

You need the definition down to the tick and you need the buyer's operational obligations in there too, because the fastest way to kill an earnout is for the buyer to change the split schedule in month two and then blame the seller when four people leave.

Naive question in a thread above my level: if the agents can leave on 30 days, why doesn't the buyer just skip the purchase and recruit the same agents directly? What does the seller have that you can't get by making better offers to the roster?

@tally the roster doesn't know you and the seller does. You're buying the introduction and the reason to say yes. That's real, it's just worth a lot less than the seller thinks it is.

@tally that's not naive, it's the whole discipline. Half the small brokerage acquisitions I've heard described were recruiting campaigns with a purchase price attached, and the buyer would have done better with a year of lunches and a better cap. The exception is when the seller is the reason the producers are there and the seller stays. Then you're buying a person and you should structure it like hiring one.

Speaking to the comparison directly, my management contracts do survive a sale, and buyers still discount them for termination clauses and owner churn. So even the sticky version doesn't price at face. A brokerage roster with no enforceable restraint should price at a steep discount to that, and in a lot of states the non-compete question for independent contractors is genuinely unsettled, so an attorney in the state has to look at what's actually enforceable before you put weight on it.

The piece I'd underwrite hardest is the fee stream. Monthly desk and tech fees keep arriving from agents who aren't producing, and that revenue is more durable than the split revenue. It's also the first thing a departing agent stops paying, so run the roster by tenure.

On the lending side of a deal like this, nobody wants to be the senior lender against contracts that terminate on notice. That tells you something. The collateral in these deals is the buyer's other assets or a personal guarantee, which means the price gets set by what the buyer will personally stand behind. That's a real constraint on the multiple regardless of which of your three frames you like.

The recruiting-platform frame is the one I'd distrust most. It means you're paying today for a process whose output you haven't seen. A machine that has been producing net productive adds for three straight years with documented cohort data, fine, price it. Most small shops that call themselves recruiting platforms have a spreadsheet of hires and no cohort survival numbers, and the hires and the survivors are very different lists.

I want the dull version, so: trailing company dollar, a small multiple, and most of it in earnout with the seller answering the phone for eighteen months. @sable, does anyone in your reading argue for paying a real multiple up front, and if so what do they think they're getting that the earnout doesn't cover?