Should I Start My Own Brokerage, or Build a Team?

Ask should I start my own brokerage and you will have a confident answer inside one click. The problem is who wrote it. The companies publishing step-by-step guides sell the coursework the license requires. The brokerages publishing comparison pages want you on their platform instead. We went looking for one guide on this question written by somebody with nothing riding on the answer, and did not find it.

Here is what the evidence actually supports. No study has ever measured whether owning a brokerage improves your income, your longevity, or your odds of still being in business in five years. The number most often used to prove that teams beat brokerages does not compare the two things. What is documented is narrower and considerably more useful: the years the license costs you, what supervision makes you personally responsible for, and the exact ceiling on what going independent can save.

Decide on those three. The income case is unmeasured, by anyone, ever.

The number everyone cites does not compare the two things

You will meet this line within a few searches: real estate teams retain an average gross margin of 61.8%, while brokerage companies retain 13.8%. Roughly four and a half times better, and it looks like it settles the question on its own.

It does not, because the two figures measure different things over differently chosen groups.

The 61.8% is a team leader's share of gross commission income after paying team members, drawn from a RealTrends survey of about 2,000 top-ranked teams. Those teams averaged $3.17 million in gross commission revenue and 10.9 years in business. That is a sample of elite survivors, and the figure sits before company overhead.

The 13.8% is company dollar retained across a brokerage's entire revenue base, after paying every affiliated agent including the ones who close nothing, and after full corporate overhead.

Put plainly: one number is what a decade-old high-producing team keeps before running costs, and the other is what an average firm of every size and model keeps after them. Both come from the same publisher, whose analyst arm sits inside a company that sells marketing and recruiting products to brokerages.

The comparison might still favor teams. Nobody has run it.

Brokers do earn more, and most of that gap is time served

There is a real gap, and it comes from a source with nothing to sell. The Bureau of Labor Statistics put the median annual wage for real estate brokers at $72,280 in May 2024, against $56,320 for sales agents. About $16,000, measured by the federal government.

Now look at what moves income inside the industry's own membership survey. In NAR's member profile, members with two years of experience or less reported a median gross income of $8,000. Members with sixteen years or more reported $88,500. Eleven to one, on tenure alone.

Those are two different surveys measuring two different populations, one reporting wages and the other self-reported gross income, and neither controls for the other. We are not going to add them together. But the direction is not subtle, and it points at the mechanism.

Because look at what the broker license costs you in time.

Every one of those requires years you must serve before you may even apply. Colorado is the strictest of the six in an easily missed way: it licenses everyone as a broker rather than a salesperson, and only active broker time counts toward the Employing Broker credential.

So the population the federal government counts as brokers is, by construction, further into the business than the population it counts as agents. Some of that $16,000 is the license. Some of it is the decade. No source we found separates them, and every guide quoting the broker premium presents the whole thing as the first.

What the license makes you personally responsible for

This is the part of the decision with no missing data, and it only runs one way.

In California, Business and Professions Code section 10177(h) allows the Department of Real Estate to discipline a broker for failure to supervise as a violation in its own right. No underlying agent misconduct has to be proven beneath it. Commissioner's Regulation 2725 sets out what you are on the hook to review and manage, including transaction documents, advertising, trust fund handling and file retention.

In Washington, RCW 18.85 requires every licensed firm to have a designated broker who holds controlling interest. That person supervises every licensee with under two years in the business and may not delegate it. Monthly trust account reconciliation sits with them. So does the firm's compliance record.

Colorado requires every active licensee, and every licensed firm with more than one broker, to carry errors and omissions coverage as a condition of licensure.

Read those together and the trade becomes visible. Right now somebody else carries this for you. Every mistake made by a colleague you did not hire, in a transaction you never saw, currently lands on a broker who is not you.

How often does it actually bite? Nobody publishes that. California, Texas and Florida each publish disciplinary actions case by case, but we could not find any of the three publishing an annual count broken out by violation type. The one certain cost in this decision is also the one whose frequency has never been measured.

The ceiling on what you save

The pitch for going independent is that you stop handing over a share of every closing. That is true, and there is a hard number on it.

Those are published caps, and the RE/MAX figure comes from a single franchise office's own plan page. There is no national schedule at a franchise. HomeSmart's own office pages contradicted each other when we checked them for our piece on commission splits, one listing a $9,000 cap and another listing none.

So the theoretical maximum a capped model can cost you is a four-figure number. That is your entire saving, before your own costs.

Now the part that matters more than the cap itself: the cap only binds if you reach it. At eXp's 80/20, reaching a $16,000 cap takes $80,000 of gross commission income in a single year. eXp's own income disclosure puts the median payment to an active agent at $13,463, across 63,410 agents paid in 2025. Most agents on a capped plan never come close to the cap, which means most agents are not paying the number they picture themselves saving.

Against whatever you are genuinely paying, here is what running a one-person firm costs every year. Some of it is published:

• Errors and omissions coverage runs roughly $700 to $815 a year, according to two companies that sell it. Every figure we could find is priced for an individual licensee. Nobody publishes what it costs to insure a firm.
• The state charges to license the entity itself. California's Department of Real Estate publishes $450 for a corporation license and $450 to renew on time. The Texas Real Estate Commission publishes $309 to register a business entity and $218 to renew. Washington's Department of Licensing publishes $304 for a firm license and $304 every two years after. We could not pin Florida, New York or Colorado to a firm-level amount on the commission's own schedule, only to individual broker renewals, so we are not printing numbers for those three. The startup line items are in what it costs to start a brokerage.
• Transaction management software starts around $32 a month per person at Dotloop. A brokerage-level product like SkySlope starts nearer $340 a month.

And then the one you cannot price at all. Most MLSs charge a broker or participant amount on top of individual agent dues, and we could not retrieve a current published schedule from any of the large ones. Their pricing sits behind a member login. The single unavoidable recurring cost of operating a brokerage is the one you cannot look up before deciding to become one.

Nobody has published this comparison end to end, in either direction, which is why we are not going to hand you a verdict dressed up as arithmetic. What the published numbers do support is narrower and more useful: your saving is capped, part of your cost is not listable, and the saving only exists at production levels most agents on those plans never reach.

The team is the third option, and it is not a small brokerage

Most people frame this as two choices. There are three, and the third is the one where the regulator does not move the liability onto you.

Teams are regulated now in a growing number of states, but almost entirely as a naming and advertising matter:

• Florida's Rule 61J2-10.026, effective July 2019, makes a team register a designated licensee responsible for advertising compliance, requires the broker to keep a monthly written record of team members, and bars team names using words like Realty, Real Estate, Brokerage or Company.
• Texas rule 535.154 requires the team name to be registered with the commission as the broker's assumed name, and to end in Team or Group.
• Oregon's HB 3137 bars "realty" and "real estate" in team names from January 2026.
• New York requires team names to carry the licensed names of the brokers or salespeople involved, and holds the supervising broker responsible for the team's advertising.
• California restricted team names implying independent brokerage status back in 2014.

Read across all of them and the same thing holds every time: the supervising broker carries the licensing-level liability, not the team leader. Florida's designated licensee has a compliance duty over advertising, which is not remotely the same as being the person a commission disciplines for failing to supervise.

That is the actual trade. A team gets you leverage over other people's production without the supervision statute attaching to your name.

What a team earns you is unmeasured in exactly the same way as everything else here. We could not find NAR or anyone else publishing what share of agents work on a team, or how team sizes are distributed. Every team number in circulation comes from surveys of top-ranked teams, and a survey of winners is not a population you can plan against.

So should I start my own brokerage? Five questions that settle it

1. Are you actually capping right now? Pull last year's statements and total what you genuinely paid your brokerage. Not the cap, the amount. If it is well under the cap, the saving you are picturing is not there to be captured.

2. Can you even apply yet? Two to four years in every state we checked, unskippable, and Colorado will not count the wrong kind of time. If the answer is "not for another two years", the real question is what you build in the meantime.

3. Are you willing to be the name on the supervision? Not "can you afford the coverage". Whether you are willing to be personally disciplinable for the conduct of people you have not hired yet, in transactions you will not see. That is the thing you are buying, and it arrives on day one whether or not any revenue does.

4. Is the firm you are picturing the common case or the rare one? There are more than 300,000 real estate firms in the United States, and when NAR last published the figure the typical one had two full-time licensees. Opening your own brokerage is not the ambitious move it gets sold as. It is the most ordinary shape of firm in the country. If your plan involves a roster of agents, you are planning the uncommon case, and recruiting is the actual project, not the license.

5. Would a team get you the same thing? If what you want is leverage on other people's production, the team reaches it without the statute. If what you want is a business you own outright, with your own brand and your own terms, the team does not, and that is a legitimate reason that needs no spreadsheet. Just do not import a financial case that nobody has ever measured.

Does owning a brokerage make you more money?

No disinterested source has ever shown that it does.

This is now the fourth time this cluster has gone looking for evidence behind a load-bearing industry claim and come back empty. We searched academic and government sources specifically for any study measuring whether brokerage ownership improves an agent's income, transaction volume, longevity or business survival. There is none. The nearest academic work is a literature review confirming that fifty years of real estate brokerage research went to pricing and market efficiency, not to whether owning one works.

The government does measure how long new businesses last, and it does not measure it for you. The Bureau of Labor Statistics publishes establishment survival only down to a category that files brokerage offices alongside equipment rental and landlords. Across all industries, about 79% of new establishments survive their first full year and roughly half are gone by year six. That is the whole country, not your business, and we are labeling it that way rather than quietly implying it is a brokerage number.

What does exist is a study by a company selling back-office accounting to brokerages, which found that fewer than two thirds of the hundred firms it looked at turned an operating profit in the first half of 2024. The selection method is undisclosed and the publisher benefits from brokerages worrying about their books. Take it as directional and nothing more.

Own it for the reason that survives the arithmetic

The financial case for opening your own brokerage has never been measured. The liability case has been written into statute in every state and is entirely one-directional. If you decide on the numbers, you are deciding on numbers somebody sold you.

The reasons that hold up are the ones a spreadsheet was never going to settle: you want the terms to be yours, you want to keep what you build, and you would rather answer for your own judgment than somebody else's. Those are good reasons. They are just not the reasons in the guides, because nobody has figured out how to sell them.

If you are going ahead, the order is license, entity, coverage, MLS participation, then the thing every recruit and every seller checks before they call you. That part is ours: live before you open, priced in the open.