Real estate team commission structures are the single most powerful retention lever a brokerage owner controls, and most owners pick one almost by accident — copying whatever their old brokerage did — then wonder why their best agents keep leaving. Compensation isn’t an HR detail. It’s the operating system of your team. It decides who you attract, who you keep, and whether your math works at 15 agents and still works at 50.
Most articles on this topic list two or three models in a paragraph each and call it a day. That’s not enough to make a decision with. This one breaks down all six models brokerages actually use, with the per-deal math, the kind of agent each one attracts and repels, the point at which each one breaks, and — the part nobody talks about — what your CRM has to be able to do to run each one without chaos.
Key Takeaways
- Six models are in real use: traditional split, capped commission, team lead override, salaried inside sales agent, mentor split, and eat-what-you-kill. Most brokerages end up running a hybrid.
- Each model attracts a different agent. A new agent wants a mentor split. A rainmaker wants a cap. Picking the wrong model for the agents you’re trying to recruit repels exactly the people you want.
- The per-deal math diverges sharply. On the same $400,000 sale, the brokerage’s cut ranges from almost nothing to nearly half, depending on the model.
- Migration is the dangerous part. Changing a commission structure badly is how you lose three top producers in a quarter. Do it on a 90-day window with grandfathering.
- Your CRM has to track commission rules at the deal level, per agent — not just at the brokerage level. Most can’t, and that’s where hybrid models fall apart.
The 6 Models, At a Glance (Comparison Matrix)
Before the deep dives, here’s the whole landscape in one place. This is the spine of any real estate team structure decision — everything below is detail on these rows.
| Model | Agent keeps | Brokerage keeps | Ideal team size | Attracts | Breaks when |
|---|---|---|---|---|---|
| Traditional split | 50–80% | 20–50% | Any | New + mid agents | Top producers demand more |
| Capped commission | 64–100% (post-cap) | Fixed annual cap | 15+ | Top producers | Cash flow gets lumpy |
| Team lead override | 50–70% minus override | Override 10–30% | 5–25 | Agents who want leads | Lead quality drops |
| Salaried ISA | Salary + bonus | All commission | 25+ (high volume) | Career support staff | Volume can’t cover salary |
| Mentor split | 40–60% (first year) | Split with mentor | Growing teams | Brand-new agents | Mentor stops mentoring |
| Eat-what-you-kill | ~100% minus fees | Desk + transaction fees | Any | Independent veterans | Agents want services |
Keep this table open as you read the rest. Each model below explains a row.
Models 1–3: Traditional Split, Capped Commission, Team Lead Override
Model 1 — Traditional split (50/50 → 80/20)
The default. The agent keeps a percentage of their commission, the brokerage keeps the rest, and the split often improves as the agent produces more.
The math on a $400,000 sale: at a 3% commission, the agent’s side is $12,000. On a 70/30 split, the agent keeps $8,400 and the brokerage keeps $3,600. Simple, predictable, easy to explain in a recruiting conversation.
Who it attracts: new and mid-level agents who value the brokerage’s brand, training, and support and aren’t yet producing enough to demand a better deal.
When it breaks: when your top producers do the math and realize they’re handing over $40,000 a year for services they’ve outgrown. That’s when they start taking recruiter calls.
What your CRM needs: the simplest case — a per-agent split percentage applied at the deal level. Almost any system handles this.
Model 2 — Capped commission (the Keller Williams model)
The agent pays the brokerage a fixed annual cap, and once they hit it, they keep 100% (or close to it) for the rest of the year.
The math: say the cap is $18,000. A top producer hits it by July, then keeps everything for the back half of the year. A part-time agent never hits it, so the brokerage effectively earns a percentage from them all year. The cap is a retention magnet for high producers precisely because it rewards them disproportionately.
Who it attracts: top producers and rainmakers. The cap is one of the most effective tools for keeping the agents most likely to be recruited away.
When it breaks: cash flow. Your highest earners stop paying you in the back half of the year, which makes the brokerage’s revenue lumpy and front-loaded. You need the financial discipline to manage that.
What your CRM needs: cap tracking per agent — a running total of brokerage-side dollars against the cap, with the split automatically flipping to 100% once the cap is met. This is where spreadsheet-run brokerages start making expensive errors.
Model 3 — Team lead override (hub-and-spoke)
The team lead generates leads — usually from paid sources or a strong sphere — and distributes them to team agents. In exchange, the lead takes an override on top of the brokerage split.
The math: the team lead might take a 20% override on the agent’s share. On that $400,000 sale where the agent’s split is $8,400, a 20% override sends $1,680 to the team lead for providing the lead, the training, and the systems. The agent trades margin for not having to generate their own business.
Who it attracts: agents who want to sell, not prospect — newer agents and those who hate lead-gen. The model lives or dies on the team lead being a genuine rainmaker.
When it breaks: when lead quality drops. If the team lead’s lead flow dries up or the leads get worse, agents feel the override as pure tax and resent it. The override has to be visibly worth it. This is also the model most dependent on disciplined lead routing rules and on tracking which lead sources actually convert — if the lead distributes garbage, the whole structure collapses.
What your CRM needs: lead routing plus two-level commission math — brokerage split and team-lead override on the same deal — and the lead-source attribution to prove the leads are worth the override.
Models 4–6: Salaried ISA, Mentor Split, Eat-What-You-Kill
Model 4 — Salaried showing assistant / inside sales agent
Instead of a commission split, you pay someone a salary plus a bonus. Usually this is an inside sales agent who qualifies leads, or a showing assistant who handles tours so producing agents can stay in front of clients.
The math: an ISA on $45,000 base plus a per-appointment or per-closing bonus. The brokerage or team lead keeps the commission those qualified leads generate. This only pencils out at real volume — you need enough deal flow that the salary is cheaper than the commission you’d otherwise pay out.
Who it attracts: people who want a real estate career with a stable paycheck — support-oriented talent who’d never survive on straight commission.
When it breaks: when volume dips and the salary outruns the commission it’s generating. Salaries don’t flex down with a slow quarter.
The W-2 wrinkle: a salaried ISA is usually an employee (W-2), not an independent contractor (1099). The IRS rules on worker classification are not optional, and getting this wrong is an expensive mistake. Talk to an accountant before you put anyone on salary.
What your CRM needs: to separate lead-qualification activity from commission entirely, and to attribute the qualified leads the ISA passes to producing agents so you can see whether the salary is paying for itself.
Model 5 — Mentor split (the new-agent model)
A brand-new agent is paired with a mentor. For the agent’s first few deals or first year, the mentor takes a share of the new agent’s commission in exchange for hands-on coaching.
The math: a first-year agent might be on a 50/50 split with the brokerage, and then split their half again with a mentor — say 70/30 in the new agent’s favor. The new agent earns less per deal but closes deals they couldn’t have closed alone, and the mentor gets paid for developing talent.
Who it attracts: brand-new agents who need a deal-by-deal safety net, and experienced agents who want to earn by teaching.
When it breaks: when the mentor stops mentoring but keeps collecting. Every mentor split needs a sunset clause — a fixed number of deals or months, after which the split ends. Without it, the new agent feels robbed the moment they’ve learned the job, and they leave.
What your CRM needs: time-boxed or deal-count-boxed split rules that automatically expire — and a clean handoff of the new agent’s onboarding, which ties directly into getting a new agent productive fast.
Model 6 — Eat-what-you-kill (100% commission with desk fee)
The agent keeps essentially all of their commission and pays the brokerage flat fees instead — a monthly desk fee plus a per-transaction fee. This is the eXp and Realty ONE Group model.
The math: the agent keeps the full $12,000 on that $400,000 sale, minus a per-transaction fee of a few hundred dollars and whatever monthly desk fee they pay. For a high-volume independent, this is far cheaper than any split. For a low-volume agent, the fixed fees can cost more than a split would have.
Who it attracts: experienced, independent agents who generate their own business and don’t want to pay for services they don’t use.
When it breaks: when agents actually want support — training, leads, brand, mentorship — and discover the flat-fee model gives them none of it. You attract self-sufficient veterans and repel everyone who needs a hand.
What your CRM needs: fee tracking rather than split math — recurring desk fees and per-transaction charges — plus enough self-service tooling that agents paying for independence actually feel independent.
How to Migrate Models Without Losing Agents
Most of the damage from commission structures doesn’t come from picking the wrong one. It comes from changing one badly.
The 90-day announcement window
Never spring a comp change on your team. Announce it at least 90 days out, in person, with the reasoning. Agents can accept a change they saw coming and understand. They revolt against one that lands in their inbox on the first of the month.
Grandfathering vs. forced migration
You usually have two choices: grandfather existing agents on their current deal and apply the new structure only to new hires, or migrate everyone. Grandfathering keeps the peace but creates a two-tier team that gets awkward fast. Forced migration is cleaner long-term but risks an exodus if you don’t handle the conversations well. For your top three producers, consider grandfathering even in a forced migration — losing one of them costs more than the structure saves.
The three conversations every affected agent needs
For anyone whose pay changes, have three separate conversations: the why (the business reason, honestly), the what (their specific new numbers, in writing, on a real deal example), and the future (what they get in return — better leads, more support, a path to a cap). Skip any of the three and the agent fills the gap with the worst-case story.
FAQ
What is the most common real estate team commission split? For traditional brokerages, a 70/30 split — agent keeps 70%, brokerage keeps 30% — is the most common starting point in the US. Top producers usually negotiate to 80/20 or move to a capped model where they pay a fixed annual cap and keep 100% after. Eat-what-you-kill brokerages charge desk and transaction fees instead of taking a split.
How do real estate team lead overrides work? In a hub-and-spoke team, the team lead generates and distributes leads, and the producing agent gets their normal split (50–70% of their side). The team lead then takes an additional override of 10–30% of the agent’s share, on top of any brokerage split, to cover the leads, training, and systems they provide.
When should I switch from a traditional split to a capped model? Switch when your top producers start asking about caps or getting recruited by brokerages that offer them. The cap is a retention tool that rewards your top 20% disproportionately and costs you almost nothing on the bottom 60% who never hit it. Most brokerages make the move around 15 active agents.
Can one brokerage run multiple commission structures? Yes, and most do — a mentor split for first-year agents, a traditional split for the middle, a cap for top producers, maybe a salaried ISA for lead qualification. The catch is operational: you need real estate brokerage software that tracks per-agent commission rules at the deal level, not just one brokerage-wide split. That’s where hybrid models usually break.
Conclusion — Pick the Model Your Team Can Actually Run
The best commission structure isn’t the one that looks smartest on a whiteboard. It’s the one your team can actually run, deal after deal, without errors, arguments, or a spreadsheet that nobody trusts. A brilliant hybrid model you can’t administer is worse than a simple split you can.
Sky Agent tracks every model in this article — splits, caps, overrides, salaried roles, mentor splits, and flat-fee structures — with per-agent commission rules applied at the deal level, plus the team visibility a broker needs to see the whole pipeline. See how it runs your structure in a 20-minute demo.