A small Australian cafe or restaurant usually sells for 1.5 to 2.5 times seller's discretionary earnings, and Skai Solutions found the test that decides it: buyers want rent under about 10% of revenue, which the ATO's own data says most small venues miss.
There are three ways the number gets arrived at, they produce very different answers for the same venue, and only one of them is the one a buyer will actually use. Below is each approach, what it is good for, where it misleads, and the one structural problem with Australian small venues that no multiple can argue around.
| Approach | Typical range | Best for | Where it misleads |
|---|---|---|---|
| Earnings multiple, on seller's discretionary earnings | 1.5 to 2.5 times | Almost every owner-operated venue. This is what buyers use. | Only as good as the earnings figure, and most owners calculate that wrong. |
| Revenue multiple | 0.3 to 0.6 times turnover | A fast sanity check across similar venues. | Ignores margin entirely. Flatters a busy venue that makes nothing. |
| Asset value | Second-hand equipment, minus removal | A floor price when there is no lease and no profit. | Ignores a profitable business. It is the worst number you can accept. |
The ranges above are Australian market ranges published by Miro Capital, a Perth advisory firm, in a piece dated 15 May 2026. It is worth naming the interest: advisory and broking firms make their money on transactions, so read their multiples as the market they observe rather than as independent measurement, and read the direction of any rounding accordingly.
Seller's discretionary earnings is the money the business produces for one working owner. It is built up, not read off a return:
The same May 2026 guidance works an example: A$45,000 net profit, A$80,000 of owner drawings, A$18,000 depreciation and A$12,000 of one-offs gives SDE of A$155,000, which at two times is A$310,000.
Two traps, and they cost real money. The first is adding back a market manager's salary rather than what you actually drew: if you take A$40,000 and replacing you would cost A$75,000, only the A$40,000 goes back, because the business can only support what it actually pays. The second is that unrecorded cash cannot be sold. A buyer's accountant reads the BAS and the GST returns, and anything that never hit the point of sale is not part of the price however real it was.
Revenue multiples exist because they are easy. Miro Capital quotes 0.3 to 0.6 times annual turnover and then says directly not to anchor on it, giving the case of a venue doing A$1.2 million on 8% net margins with an owner working six days a week, which is not worth A$600,000.
Our own restaurant profit margin page explains why this happens so reliably in Australia. On ATO 2023-24 small business benchmarks, average total expenses run 84% of turnover for a restaurant billing A$65,000 to A$500,000, 88% from A$500,001 to A$2 million and 91% above that. Subtract and the implied pre-tax margin falls from 16% to 12% to 9% as venues get bigger, and for a sole trader that is before the owner is paid at all. So a revenue multiple applied to a large venue is being applied to the thinnest margins in the sector.
This is the floor, not a valuation, and you only end up here when there is no transferable lease or no verifiable profit. Published advisory guidance lists what buyers explicitly do not pay for: fit-out cost, brand without measurable repeat visits, supplier and roaster relationships, and the years invested. What is left is equipment at second-hand value, and our commercial kitchen equipment cost page shows how quickly that falls.
One consequence worth planning around: because fit-out is not recoverable on exit, the money spent on the room is spent for the trading years, not for the sale. That is an argument for spending it early, not late.
In hospitality the lease is the asset. Without the right to occupy the space there is no income stream, so there is nothing to buy. The buyer criteria published in May 2026 are specific: three to five years remaining as a minimum and ideally five plus options, rent below about 10% of revenue with some tolerance to 12% or 13% for premium sites, a landlord who will assign, and no demolition or redevelopment clause.
The same source gives the clearest illustration of what happens when the lease is short. A fourteen year old suburban cafe netting A$95,000 of SDE, well known locally, with eighteen months left and no options: the owner wanted A$650,000 and it sold for A$140,000. That is 1.47 times SDE against a 1.5 to 2.5 band, and the discount was the lease, not the coffee.
On its own published figures, a great many do, and this is the single most useful thing on this page because nobody puts the two sources side by side.
| Business and turnover band | ATO rent as share of turnover, 2023-24 | Under the 10% buyers want? |
|---|---|---|
| Restaurant, A$65,000 to A$500,000 | 11% to 17% | No, the whole range is above it |
| Restaurant, A$500,001 to A$2,000,000 | 8% to 12% | Partly, the upper half is not |
| Restaurant, over A$2,000,000 | 6% to 9% | Yes |
| Coffee shop, A$65,000 to A$250,000 | 10% to 17% | Barely at the bottom, no above it |
| Coffee shop, A$250,001 to A$600,000 | 8% to 14% | Partly |
| Coffee shop, over A$600,000 | 6% to 10% | Yes |
The rent shares are the ATO's, from lodged 2023-24 returns; the comparison against the 10% buyer threshold is ours. The pattern is stark: the smallest venues, which are the ones most likely to be sold by an owner-operator, sit entirely or mostly above the level a buyer treats as acceptable. Rent as a share of turnover falls steadily as venues get larger, so the venue most likely to be for sale is the one least likely to clear the test.
Which points at the one lever available. Rent is fixed by the lease, so the only way to move rent as a share of revenue is to move revenue. Our restaurant occupancy cost page works through that in detail, including why there is no published New Zealand equivalent of these bands.
One line item almost never appears in a hospitality sale discussion, and in Victoria the statute is explicit about it. If an employee stays with the business after it is sold or ownership transfers, their years of service still count towards long service leave accrual, and the new owner becomes responsible for any existing entitlement.
In an industry where the kitchen hand and the Tuesday closer often outlast three owners, that is real money crossing at settlement. The accrual rate is the same 0.8667 weeks a year in New South Wales, Queensland and Victoria, so one employee with ten years of service carries about 8.6667 weeks of paid leave, and a long serving casual on 25 hours a week for ten years works out at roughly 216 hours. Skai Solutions has set the arithmetic out on long service leave in hospitality, including the Queensland formula for casuals and the two figures we could not source.
Two consequences for the number at the top of this page. A buyer who keeps the team is buying the accrued entitlement with them, so it belongs in the adjustment at settlement rather than in a footnote. And a seller who has never provisioned for it has been running on an understated cost base, which means the earnings the multiple gets applied to are flattered. Check it before you agree a multiple, not after.
Use the earnings multiple, because that is what the buyer will use. Then check it against revenue as a rough cross-reference, and keep asset value in mind only as the floor you refuse to go below.
Then do the two things that actually move the number, both of which take about a year. Secure the lease: talk to the landlord twelve to eighteen months out, while you are still a reliable tenant with something to trade. And reduce how much of the business is you. A venue where the owner is the head chef, the buyer, the bookkeeper and the person who knows every regular is discounted for the risk that the trade leaves with the owner. Rostered systems, a manager in place, and documented supplier agreements are what turn a job into a business a stranger can buy.
None of this is financial or legal advice and Skai Solutions is neither a broker nor a valuer. It is what the published Australian sources say, gathered in one place so you can go into a conversation with your accountant already knowing which number matters.
Only in one place, and the connection is narrower than a vendor would like. Buyers pay for earnings that do not depend on the owner and for trade that is verifiable. A phone answered by the owner's mobile is both of those problems at once: the bookings live in one person's head, and nothing about them is in the accounts. Every order captured through a system leaves a record, and a record is what a buyer's accountant can price.
That is a marginal improvement to a sale, not a reason to buy anything. Its real effect is on the numerator of the rent test above, which is revenue. See how to answer the phone at a restaurant for the version that costs nothing, and AI receptionist for the version that does not.
Most small Australian cafes and restaurants sell for 1.5 to 2.5 times seller's discretionary earnings. Miro Capital, an Australian advisory firm, published a range of A$100,000 to A$450,000 for a typical cafe on 15 May 2026. The multiple is set less by your food than by your lease, your reliance on the owner, and whether the earnings are verifiable.
Net profit plus the owner's wage and drawings, plus depreciation, plus genuine one-off costs. It is the money the business generates for one working owner. It is not your net profit, and it is not your turnover, which is where most valuations go wrong before the multiple is even applied.
Brokers do quote revenue multiples, commonly 0.3 to 0.6 times annual turnover, and they warn against relying on them. A venue turning over A$1.2 million on thin margins with the owner working six days is not worth A$600,000. Use revenue as a sanity check on an earnings figure, never as the valuation.
Because in hospitality the lease is the asset rather than a cost. Without a transferable lease with enough term left, there is no income stream to buy. Australian advisory guidance published in May 2026 describes three to five years remaining as a minimum, ideally five plus options, and rent under about 10% of revenue.
Very little. Advisory guidance is blunt about it: buyers do not pay for fit-out cost, brand without measurable repeat trade, supplier relationships, or the years you put in. Equipment has a second-hand value and that value falls fast. What gets paid for is future earnings plus the right to occupy the space.
Twelve to eighteen months before you want to be out, mainly so the lease can be extended or an option exercised while you are still the sitting tenant, and so a year of clean, verifiable trading sits in the accounts. Both of those take time you cannot buy back later.