What your split is really buying

Is this you

You are negotiating a percentage and you have no idea whether the number is fair, because nobody publishes theirs.

It shows up in three ways. A broker you want asks for eighty and you do not know whether that is bold or normal. A broker you have starts asking for more and you cannot tell whether they have an offer or a grievance. Or you are designing the seat from scratch and every number you pick feels arbitrary, because it is.

Underneath all three is the same problem. You are treating the split as a price when it is a trade, and you have not written down what you are giving.

The market is no help. Across 991 finance advertisements over six months, the number naming a commission split at all sits between barely a dozen and about two dozen, depending on which of two honest implementations of the same pattern you run. Either way it is almost nobody. Just over half named a salary. So the one figure a broker will decide on is the one the market has agreed to keep quiet about, and you are negotiating it blind against someone who has probably had the conversation more recently than you.

There is a reason the number stays quiet, and it is not only competitive nerves. One Sydney principal told us how he handles it. 'So I don't really tell people what they are until I talk to them. I know what they can bring. If someone, if I think someone's super going to be really, really good, I'll give them a lot higher.' He is not withholding a rate card so much as pricing the person in front of him. Which means the number a broker hears first is a read on that broker, and if that is how you set yours, there is nothing written down for them to negotiate against and nothing written down for you either.

BD Machine ad index, 991 advertisements, 23 Feb to 20 Aug 2026

The split is priced off who carries the cost and the risk

Most of the real structures we have seen price the same thing, who is carrying the cost and the risk. Who found the client is the largest part of that, which is why the number moves when the lead flow moves. It is not the whole of it.

A principal designing his first salaried broker role set the rule out better than we would. 'That's the thing when in broking, if you come in and you have full security and everything's paid for, the percentage of commission that you get is not going to be as high because you literally put in none of your own money. Right. But if you're having to pay for things [...] Then you can usually get like a 50, 50 commission when it's commission only and then like a 70, 30 when it's like self gen. But if he's on the salary and he's got everything provided for him, I. I think 50, 50 doesn't make sense.'

That is the shape of it. How wide the fork is, though, is a decision, and plenty of businesses make it far narrower than fifty and seventy. One broker is on a sixty thousand base, twenty per cent of commission up to thirty thousand, twenty five above that, and thirty per cent on his own referrals. That is a premium of five to ten points for finding the client himself, not twenty. Whatever his principal meant by it, a five point premium tells a broker that hunting is barely worth doing.

The other half of the trade is what your share pays for, and it is not only marketing. The principal quoted in section 02 lists the leads first and then keeps going, to an upfront person, a settlements person and a credit resource. A founding partner of a regional finance firm described his own fifty fifty the same way, as 'the balance of the 50 funds, the OPS, your support stuff and everything else that sits in the background'. In both businesses a large part of the house share pays for work that happens after the broker has already won the client.

So a low split is not mean and a high split is not generous. A fifty per cent split with real lead flow can pay a broker more than eighty per cent of what they can find alone, and both of you can check that in an afternoon. What makes the negotiation go badly is doing it in percentages instead of in dollars, so do yours in dollars.

Percentages hide whole layers. One writer describes his arrangement as seventy thirty. 'We split our commission 70/30 with mortgage brokers. So if I settle $100 grand, the business takes $30,000'. He is an employee of the thirty, not a partner in the seventy, so the number he is actually paid sits somewhere inside that thirty. If your broker cannot say what their percentage is a percentage of, neither can you.

Recorded conversations with principals and brokers, 2025 to 2026, verified against the transcripts and anonymised

What the market will and will not tell you

Almost none

of 991 advertisements mentioned a split

Over six months, across Australia. The number a broker decides on is the one the market has agreed not to publish, so every negotiation starts from nothing.

BD Machine ad index, 991 ads, 23 Feb to 20 Aug 2026. Counted on an explicit split or profit-share phrase. Two honest implementations of that pattern disagree by about a factor of two and the tighter of them falls below the publish gate, so this is described rather than counted

153

advertisements offered the leads instead

Businesses advertise the lead flow many times more often than they advertise the split. They are selling the engine, which is the half of the trade that is easy to describe.

BD Machine ad index, 991 ads. 105 of the 153 also named a salary, against 53% across the window

524 of 991

named a salary at all

Just over half. So a broker comparing two businesses is usually comparing one number against no numbers, and filling the gap with whatever they heard last.

BD Machine ad index, 991 ads

74% vs 49%

recruiter against business

When a recruiter placed the advertisement, 121 of 164 named a number. When the business placed it itself, 403 of 827 did. The people who do this for a living have already worked out that saying the number wins the candidate.

BD Machine ad index, company-name pattern rather than a verified classification

The same volume, three ways

Upfront commission runs about six and a half thousand dollars for every million settled, the bottom of the MFAA's 0.65% to 0.70% range. That is the number this whole conversation is built on.

Read across the row you can actually deliver, not the one you would like to. A broker on fifty per cent of house leads at two million a month earns more than a broker on eighty per cent who can only find one million on their own, and the second one is working far harder for it.

Then take something off your row, because the percentage is rarely the first thing that happens to the money. Most of the priced structures we have been shown have something standing in front of it. One asset finance broker on a base has seven thousand withheld gross against admin costs, is charged a two hundred and fifty dollar admin fee on every lead, and only then earns forty per cent of the first twenty two thousand of commission and sixty per cent above it. He is not on forty per cent, and he is not on sixty either. Whatever stands in front of your percentage, say it out loud, because he will find it anyway.

One broker ran that exercise on an offer in front of him and reached twenty eight per cent before the business had done the sum at all. 'It sounds like 50. It's just going to be around 28 according to my calculation though. Because they're gonna. Minus the base as well.' The headline rate and the realised rate are two different numbers, and your candidate works out the second one whether or not you do.

Trail sits on top, in the order of a hundred and twenty five dollars a month for every million written, and two things about it go missing from most conversations. The first is who keeps it, which section 07 comes back to, because it decides whether two offers quoting the same fifty are the same offer at all. The second is that trail does not simply compound. It compounds while the book is being fed and it shrinks when it is not. One sole operator watched his own fall from about six thousand a month to about one thousand as clients refinanced away.

None of this is the whole package. It is the part that gets argued about.

FinTalent FY26/27 Salary Guide for the upfront and trail rates. The table is arithmetic on the upfront rate. The structures described around it are from recorded conversations with brokers, anonymised

The fork

The two bands are FinTalent's read across a small number of our own client engagements rather than anything a principal said or a published range, and the conversations behind this guide show forks far narrower than the bands suggest. The quotes around them are from recorded conversations with principals and brokers, 2025 to 2026, anonymised

Work out your own number

An afternoon with a calculator settles what a year of arguing will not.

Most of what follows is not about the percentage.

1. Count the leads you hand over

   Not enquiries. Leads a broker can call, with a name and a reason. Count last month, not a good month.

   If the honest answer is none, you are in the right-hand column and your split should say so. Businesses get this wrong in both directions, and paying a low split while supplying nothing is the fastest way to lose a good broker.

2. Price the whole engine

   Add up what those leads cost you. Marketing, the referral relationships you maintain, the brand, the person who books the appointments. Then keep going, because the engine does not stop when the lead lands. The upfront person, the settlements person and the credit resource all come out of your half, and so do employment on-costs.

   Divide by the number of leads and you have something to compare against. One founder can afford a higher split than the businesses around her because her marketing cost is her own time and she is not paying out twenty per cent referral fees. If your engine costs you less than hers, that is your argument, and most principals never make it.

3. Run the row, then run it net

   Take the table above. Work out what the broker earns on your leads at your split, and what they would earn at eighty per cent finding their own.

   Then do it again after the gate, the recovery and the deductions, because that is the row the broker will run. The broker who worked his own offer down to twenty eight per cent in section 05 was reading an offer that said fifty. Another, a senior lender on about two hundred and forty thousand all in, worked out that writing enough volume to sit near the top hundred brokers in the country would land him roughly on the package he already had, and could not make it stack up.

4. Say what the percentage covers

   Gross upfront, or what is left after expenses? Credit files, title searches, admin fees, lender fees. One broker pays one per cent of the net amount financed on loans written to particular lenders, so a fifty thousand dollar loan costs him five hundred dollars out of the two and a half to three thousand he earns on it, on top of whatever the split takes.

   Then say who wears a clawback. A broker choosing between salaried offers told us clawback and flexibility were his deciding issues rather than the percentage, and he raised it himself at the end of the call because he still did not have an answer from anyone. If you cannot say in one sentence who carries a clawback, the candidate will find the hole before you do.

5. Check the gate against the seat

   If commission starts at a threshold, three questions. What is it measured in, settled volume or revenue? Gross or net? And has anybody in that seat ever cleared it?

   One broker sits in a seat whose bonus he cannot reach, because the self-managed super fund deals he is fed are capped in size by their nature, so no amount of effort lifts the number. 'You can't have a $5 million loan in a self managed super fund.' Later in the same conversation he gave us the record of the role. 'In two years there has been four people worked in this role and none of them have ever met the target.'

   In a different business, an employed broker described a structure that keeps everything until two hundred and fifty thousand dollars of revenue and pays twenty five per cent above it, on referral flow patchy enough that he can go months without writing anything. Ask whether anyone in your seat has ever cleared your gate, because a bonus nobody reaches is decoration, and the broker has already priced it at zero while you are still counting it as part of the package.

6. Is the base a wage or an advance

   The cheapest fix in this document, and almost nobody makes it. The broker who worked his offer down to twenty eight per cent asked the first question about a base plus split moments later in the same conversation. 'So my only point is like someone's gonna be very quiet and what if I'm just hitting like 15, 20 grand? Is that something I'm still gonna get my base or are they gonna deduct from. From what?'

   If the base is recovered out of commission it is a draw, and the broker's real earnings are the table's figure minus the base rather than on top of it. One broker described a structure where she had to write a set number every year simply to cover her own salary before any bonus began. Say which one yours is, in the advertisement, and you will be the only one who does.

7. Say who keeps the trail

   Trail is the compounding half of a broker's income and it gets negotiated far less than the upfront. On a business-generated lead plenty of firms keep all of it, which means two offers quoting the same fifty per cent are not the same offer. That is our own read of the market rather than a principal's account of it, and so is the figure that follows. On one engagement we negotiated fifty fifty across upfront and trail on business-generated leads and asked for seventy thirty on self-generated business, which is a materially different seat from fifty on the upfront alone.

   Then say what happens when they leave. One broker was a few loans into a mentoring arrangement when he worked out the trail stayed behind. 'So it comes back to I'm actually building someone else's business, not my business.' Another stayed more than a decade and described what held her. 'You still can't take your trail. And so you were stuck.'

8. Write the fork down first

   Decide the house-lead number and the self-generated number, and say both. Every business we have seen do this well runs two rates, not one.

   Almost nobody does, and the Sydney principal in section 01 told us why he does not. He sets the number after he has met the person, so there is nothing written down until it is too late to be useful. Write it first and you end the negotiation early and in your favour, because a broker who understands the trade stops arguing about the percentage.

What people do instead

1. Matching a number they heard. A broker says another business offered eighty, and the principal either matches it or loses the candidate, without either of them asking what came with the eighty. Almost always, nothing did. The reverse costs just as much. A high rate sitting behind a gate the broker cannot reach loses you the candidate and your credibility in the same conversation, and one told us exactly how he reads it. 'So that way, you know, when I go in, it's 85 and they're like, ah, nah, you need to do 20 deals a month to get that. I was like, whoa, I can't reach that.'
2. Running one rate for both kinds of deal. It quietly punishes whichever broker is doing the thing you actually want. If house leads and self-generated pay the same, you have made the engine free and the hunting worthless. One broker told us he does not focus on self-generating at the moment, because nothing in the structure pays him a trail or any other incentive for doing it, and he would rather spend that time there than on settlements and document requests. If that is your business, you are short of self-generated flow because you are not paying for it.
3. Designing the split before designing the seat. The percentage is the last decision, not the first. What you supply decides it, and most businesses pick the number first and then find out what they can afford to give. It also decides who applies. A broker support manager watching her own firm hire said it plainly. 'They won't have a client database because of the splits and everything.' Your split is a recruitment filter before it is a payment structure, and it filters whether you meant it to or not.
4. Quoting the top tier as though it were the rate. A tiered structure does not pay its headline. Take the asset finance broker whose package section 05 itemises, on forty per cent of the first twenty two thousand of commission and sixty per cent above it. A forty thousand dollar commission month pays him eight thousand eight hundred on the first slice and ten thousand eight hundred on the second, a shade under fifty per cent rather than sixty. Worse, the moment a threshold exists it schedules the work as well as pricing it. He told us how he plays his own tier. He gets his smaller deals through up to the threshold and holds the bigger ones until after he has hit it, 'so I can get the biggest percentage on it'. His clients are being sorted into months. His principal has no idea.
5. Capping the thing you are asking someone to chase. One broker is on a hundred and twenty thousand base with commission capped at six thousand, which is the whole of his upside for writing more. If a cap is the only way you can make the top band affordable, look at what could fund it instead. One principal who steps his split upward as monthly commission rises is comfortable at the top of it because the trail book underneath keeps growing while his big writers write.
6. Paying a split nobody can see. One lead-generation officer paid on settled deals spends the end of every month working through fifty odd leads by hand, in a spreadsheet, to establish what actually settled and therefore what he is owed. If that is your business, what you have built is a monthly argument, and the cheapest improvement you can make to any commission structure is showing the person what they have already earned.