Splits, caps, or fees: what actually survives if agent counts shrink and commissions don't?

The cap is the only one of the three I'd sign without arguing, at least going by what I've dealt with as an agent. A split feels like a tax that never ends and a desk fee feels like rent I owe whether or not I ate that month. The cap has an end date on the calendar and I can see it.

Depends entirely on whether you're recruiting producers or growing them, which plumb's setup leaves open.

If your ten agents are experienced people doing 15 to 30 sides, a fee model wins and you don't need much infrastructure because they came with it. If your ten are two years in and doing five sides, you need to fund training and file review out of your revenue, and a fee model gives you $89 a month to do it with. That's a split shop whether you like the optics or not.

The mistake I see is picking the model that recruits well and then discovering you've recruited people the model can't afford to support.

What's a desk fee actually paying for if the brokerage has no desks anymore? Genuine question, half the firms near me are fully remote.

rigging, in the shops I've been inside it's paying for the broker of record, E&O coverage, the tech stack seats, and file review. The desk is a historical artifact.

plumb, the variable your setup doesn't include is churn cost. A fee model has almost no switching cost for the agent, so your roster turns over faster and you eat recruiting cost every year. A cap has a built-in reason to stay through December because leaving in October means you paid the cap somewhere and start over. That's worth real money and nobody prices it.

Coming at this from having sold services to contractors for years, the flat fee model has one failure that never shows up in the spreadsheet. When your revenue is a subscription, your incentive to help an agent close more deals goes to roughly zero. You get paid the same. Over three years that shows up as a firm that does nothing for anyone, and then price is the only thing you compete on, and someone charges $199.

Voted hybrid, because it's what my own business converged on after five years of trying the pure versions.

Low base split, then charge separately for the things that actually cost you money to deliver. Transaction coordination at a per file rate. A paid training program that new agents are required to buy in their first year. Lead generation as an opt-in with a referral fee attached. It's messier to explain and every agent argues about one line item, but the revenue lines up with the cost lines, which is the only property of a pricing model that matters when volume drops.

The pure fee model made me hate my most demanding clients. The pure percentage model made my best clients hate me.

solder's point is the one I'd weigh heaviest. I've watched the same dynamic in property management at scale, where a flat per door fee means the manager's best move is to do less per door. The brokerage version is a broker of record who signs off on files and never picks up the phone.

Though thatch's version has its own problem. Every separate charge is a negotiation, and with ten agents you'll have ten different arrangements inside two years because you said yes to the producer who asked. Then you can't tell what any of it earns.