How an earn-out works when you sell a business
An earn-out is the part of the sale price a buyer pays only if the business performs, after you have already handed it over. It is not a bonus and it is not goodwill — it is your money, held back, and made conditional on results you will no longer control. That single sentence is the whole subject: everything that follows is about how the target gets defined, who that definition favours, and what a seller can do about it before signing rather than after.
What an earn-out is, and why a buyer asks for one
When a business changes hands, the price does not have to arrive in one piece. A typical structure splits it three ways:
- Cash at completion. Paid on the day, unconditional. This is the only part of the price that is genuinely certain.
- Deferred consideration. A fixed sum paid later, on a fixed date. The amount is already agreed; only the timing is delayed. The risk here is that the buyer can still pay, not whether they have to.
- Earn-out. A sum paid only if the business hits agreed targets over an agreed period after completion. The amount is not agreed — only the formula that will produce it is.
Sellers routinely collapse the last two together, and it is an expensive mistake. Deferred consideration is a debt: the buyer owes it whatever happens. An earn-out is a contingency: the buyer may end up owing nothing at all, entirely lawfully, and still be able to say the deal was done at the headline figure.
So why does a buyer want one? Not, usually, to cheat anybody. A buyer wants an earn-out because of a genuine information problem that no amount of due diligence fully solves.
The gap the buyer is trying to close
A buyer is paying today for profits that arrive tomorrow. Everything they can inspect — the accounts, the contracts, the customer list, the management information — describes a business that was run by you. The thing they are buying is a business that will be run without you. Between those two businesses sits a gap, and the size of the gap depends on things that are hard to see from the outside:
- Owner dependence. If the biggest customers deal with you personally, the buyer cannot tell from a contract whether those customers stay. The contract has your name on it; the relationship does not transfer with the paperwork. This is the same dependence problem that drives the whole discussion in how to value a small business.
- Recent growth that may or may not repeat. A business whose profits jumped last year is worth more if the jump is structural and worth much less if it was one unusual contract. The seller believes the first. The buyer suspects the second. Neither can prove it.
- Concentration. If one customer is a third of revenue, the buyer is not really buying a business, they are buying a relationship with a risk attached, and they will not pay a clean multiple for it.
- A pipeline nobody can verify. Work that is likely but not signed is worth something to a seller and nothing to a buyer, because the buyer cannot audit optimism.
Faced with that gap, a buyer has three options. They can pay the seller’s price and carry the risk. They can discount the price and carry less. Or they can propose an earn-out, which says: I will pay your number if your number turns out to be true.
Put that way it sounds fair, and often it is. An earn-out is the honest way to settle an argument where both sides believe different futures and neither can prove theirs. It is also, in practice, the mechanism by which a seller ends up financing part of their own exit — because until the earn-out period ends, the seller is still exposed to the business, but no longer in charge of it.
The thing sellers get wrong first
The single most common error is treating the headline number as the price. A buyer offers “£1.6 million” and the seller hears £1.6 million. What was actually offered was, say, £1.0 million at completion and up to £600,000 more if a set of conditions are met over two years.
Those are not the same offer, and a competing bid of a flat £1.25 million in cash may well be the better one. Comparing offers means comparing the certain parts first, then asking separately what the conditional part is really worth given who controls the conditions. A useful discipline when you are weighing bids — covered more generally in how to sell a business — is to write every offer as two numbers: what you get if things go badly, and what you get if things go perfectly. The gap between those two numbers is the part of the deal you need to read properly.
How the target gets defined, and why that choice decides who wins
Every earn-out reduces to one formula. Learn this shape and the rest of the negotiation becomes legible:
earn-out payment = maximum earn-out × (actual result ÷ target result), set to nil below the threshold and capped at the maximum
There are therefore five variables, and all five are negotiable:
- The measure — revenue, gross profit, EBITDA, or something narrower like retained customers or units shipped.
- The target — the number that measure has to reach.
- The maximum — how much money is at stake.
- The threshold — the floor below which nothing is paid at all.
- The period — how long the clock runs, and whether each year is measured on its own or the whole period is added together.
Sellers almost always negotiate the second and third and ignore the other three. The first one — the measure — is the one that decides who wins.
Revenue, gross profit, EBITDA: three different worlds
| Measure | What it is | Who it favours, and why |
|---|---|---|
| Revenue | Money invoiced, before any cost is taken off. | The seller. It is the hardest number to manipulate and the easiest to verify — it comes straight off the sales ledger. The buyer’s cost decisions, overheads, group charges and investment choices cannot touch it. Buyers resist it precisely for that reason: a seller can hit a revenue target on work that loses money. |
| Gross profit | Revenue less direct costs of delivering it. | A genuine middle. It stops revenue being bought with discounts, because a sale at a crushed margin barely moves it. It still sits above the overhead line, so most of what a new owner changes — head office costs, management salaries, systems spend — does not reach it. If you are being pushed off revenue, this is the line to push back to. |
| EBITDA | Earnings before interest, tax, depreciation and amortisation — profit after overheads. | The buyer. Everything the new owner decides to spend lands above this line. A hire you would not have made, a rebrand, a systems migration, an allocated share of group costs — each one reduces EBITDA and each one reduces your payment, whether or not it was a good decision. |
| Narrow operational measures | Retained customers, contracts renewed, units shipped, headcount kept. | Depends entirely on drafting. These can be excellent for a seller because they are objective and hard to argue about. They can also be brutal, because a single lost contract can cross a threshold and take the whole payment with it. |
The further down that table you go, the more of the buyer’s own conduct sits between your work and your cheque. That is the entire reason the choice of measure matters more than the size of the target. A generous EBITDA target you do not control is worth less than a modest revenue target you do.
Which profit figure is even being discussed is its own trap in a small business, because owner-manager accounts and buyer accounts rarely use the same line. If the earn-out is written against “profit” without defining it, you have not agreed anything. The difference between the seller’s view of earnings and the acquirer’s is set out in SDE vs EBITDA, and it matters here twice over: once when the headline multiple is set, and again every year the earn-out is measured.
Threshold, cap and the shape in between
Three shapes appear again and again, and they behave very differently:
- The cliff. All or nothing at the target. Hit 100%, get everything; reach 99%, get nothing. Simple to draft and savage in practice, because it makes a rounding error worth six figures.
- The straight-line slide. You are paid in proportion to performance, usually between a threshold and a cap. Miss by 4% and you lose 4%. This is the fair shape and it is worth conceding a lower maximum to get it.
- The stepped scale. Bands — nothing below 85%, half between 85% and 95%, full above 95%. A compromise, but each band edge is a small cliff, so a seller should check where their realistic outcome sits relative to the nearest step.
The cap deserves a moment too. Most earn-outs cap the upside at the maximum, which means the seller carries the downside in full and shares none of the upside beyond the target. If the buyer wants a target set at an ambitious growth figure, the natural counter is an uncapped slide above it, or a higher rate above 100%. A buyer who wants you to accept the risk of a stretching target but refuses to pay for exceeding it is asking for one side of a two-sided bargain.
Worked example: one trading year, three definitions, three different cheques
Here is why the measure decides the deal. Take one business and one real trading year, and pay the earn-out three different ways.
The business in the year before the sale:
| Line | Sum | Amount |
|---|---|---|
| Revenue | £1,800,000 | |
| Direct costs (60% of revenue) | £1,800,000 × 60% | £1,080,000 |
| Gross profit (40% margin) | £1,800,000 − £1,080,000 | £720,000 |
| Overheads | £320,000 | |
| EBITDA | £720,000 − £320,000 | £400,000 |
The deal: £1,600,000 at completion, plus an earn-out of up to £250,000 measured over the first year after the sale. The target is simply that the chosen measure holds at the pre-sale level — no growth required. Payment slides in proportion to performance, with nothing paid below a threshold of 80% of target and nothing extra above 100%.
What actually happened in year one
The buyer wanted volume, so they discounted to win two larger accounts and hired a sales manager to service them. Revenue rose 10%, margin fell from 40% to 35%, and overheads rose by £13,000:
| Line | Sum | Amount |
|---|---|---|
| Revenue | £1,800,000 × 1.10 | £1,980,000 |
| Gross profit at 35% | £1,980,000 × 35% | £693,000 |
| Overheads | £320,000 + £13,000 | £333,000 |
| EBITDA | £693,000 − £333,000 | £360,000 |
Same year. Same business. Same buyer. Now run the formula three times:
| Measure | Target | Actual | Percentage of target | Payment |
|---|---|---|---|---|
| Revenue | £1,800,000 | £1,980,000 | £1,980,000 ÷ £1,800,000 = 110% | Capped at the maximum: £250,000 |
| Gross profit | £720,000 | £693,000 | £693,000 ÷ £720,000 = 96.25% | £250,000 × 96.25% = £240,625 |
| EBITDA | £400,000 | £360,000 | £360,000 ÷ £400,000 = 90% | £250,000 × 90% = £225,000 |
A spread of £25,000 on identical trading, decided by nothing except which line of the profit and loss account the lawyers wrote into the schedule. Note also the direction of travel: on revenue the seller is rewarded for a year in which the business got less profitable, and on EBITDA the seller is punished for a growth decision the buyer made. Neither measure is fair in isolation. Gross profit — the middle row — tracks what actually happened most honestly, which is why it is so often the sensible landing point.
Now add one line nobody discussed
Suppose the buyer is a group, and the group allocates a £60,000 management charge to its new subsidiary: a share of head office finance, HR and IT. It is an entirely normal accounting practice. It may even be a fair reflection of services received. Recalculate:
| Line | Sum | Amount |
|---|---|---|
| EBITDA before group charge | £360,000 | |
| Group management charge | − £60,000 | |
| EBITDA as reported | £360,000 − £60,000 | £300,000 |
| Percentage of target | £300,000 ÷ £400,000 | 75% |
| Threshold | 80% | |
| Payment | Below the threshold | £0 |
The earn-out has gone from £250,000 to nothing. Nobody has done anything improper. The business grew. The customers stayed. One accounting entry, made by the buyer, on a line the seller never negotiated, crossed a threshold the seller agreed to without modelling it.
This is the case for doing the arithmetic before signing rather than after, and for doing it at bad numbers as well as good ones. Take whatever structure is on the table, write out the year at the plan, at 10% below the plan, and with one unexpected cost line added, and see what each one pays. If the answer swings by six figures, the structure is not yet finished.
The control problem: paid on results you no longer run
Here is the contradiction at the centre of every earn-out. The seller’s remaining money depends on how the business performs. The buyer decides how the business is run. Those two facts sit inside the same contract and they pull in opposite directions from the morning after completion.
It is worth being precise about this, because sellers tend to imagine the risk as dishonesty and it usually is not. The ordinary, good-faith decisions a new owner makes are quite enough to destroy an earn-out on their own:
- Investment. A buyer who spends £80,000 on new systems, a rebrand or a sales team is doing the right thing for a business they intend to own for a decade. Every pound of it comes off an EBITDA-measured earn-out in the one year you happen to be measured on.
- Integration. Your customers get migrated onto the group’s contracts, your product gets folded into their range, your revenue gets recorded in their system under their codes. Three months in, the thing your earn-out measures may no longer exist as a separately identifiable business at all.
- Cost allocation. The group management charge above. Also insurance, legal, audit, software licences — all now bought centrally and recharged on some reasonable basis you never saw.
- Accounting policy. Revenue recognition timing, stock valuation, capitalising versus expensing development work, bad debt provisioning, when a contract counts as delivered. Each one is a legitimate choice, and each one moves the number your payment is calculated from.
- Priority. The buyer owns several businesses. When a good lead could go to yours or to another one, and yours costs them an earn-out payment, nobody has to do anything wrong for the lead to end up elsewhere.
- Pace. A buyer who slows a project, defers a launch, or pushes a large order into the following quarter has not harmed the business. They have moved it out of your window.
The seller who stays on, which is most of them
Because of all this, earn-outs usually come with the seller staying in the business for the earn-out period. That is meant to solve the control problem. It does not, and it is important to see why.
You will have influence, not control. You report to someone now. You need approval for spending you used to authorise yourself. Decisions you would have made in a morning go to a board that meets monthly. And your position is genuinely awkward: you are simultaneously an employee who is supposed to support the new owner’s strategy, and a creditor whose payment that strategy might reduce. Every disagreement about how to run the business is also a disagreement about your money, and both of you know it.
Two consequences follow, and sellers should decide how they feel about both before they agree to a large earn-out rather than discovering it in month four:
- You may be managed out. If the contract says your earn-out lapses when you leave, your continued employment is worth real money to you and real money to whoever decides whether you stay. Any earn-out conditional on employment needs to define what happens if you are dismissed, if you resign because the role has materially changed, and if you cannot work through illness. A common landing point is that the earn-out survives unless you resign without cause or are dismissed for serious misconduct — but it has to be written down.
- You may want to leave anyway. Plenty of owners sell precisely because they have had enough. Agreeing to two more years in a business you no longer own, under somebody else’s process, in exchange for money that might not arrive, is a real cost. Price it honestly. If you would pay £100,000 to walk away on completion day, then a £150,000 earn-out requiring two more years is barely worth having.
The test worth applying before you agree to anything
Ask one question of every earn-out structure put in front of you: which of the things that determine this payment can I still affect, and which can the other side?
Write the answer in two columns. If your column contains “work hard, keep customers happy” and theirs contains “set the budget, allocate costs, choose accounting policy, decide what gets integrated, decide whether I stay”, you are not being paid for performance. You are being paid at the buyer’s discretion, in a document that does not say so. That is the point at which you either move the measure up the profit and loss account, shorten the period, shrink the earn-out and take more cash at completion, or walk. The arithmetic above tells you what each of those trades is worth; the columns tell you which one you need.
What a seller negotiates, and the second worked example
An earn-out is not something you accept or refuse. It is a set of terms, and almost all of them are moveable if you raise them before the heads of terms harden. These are the protections that do real work, roughly in order of how much money they are worth.
1. Move the measure up the profit and loss account
Everything else on this list is damage limitation compared with this. Revenue or gross profit removes most of the buyer’s discretion at a stroke. If the buyer insists on EBITDA, insist in return on a defined EBITDA: a written list of what may and may not be charged against it. Typically that means excluding group management charges and recharges, acquisition and integration costs, costs of the transaction itself, any new expenditure above an agreed threshold that you did not approve, and any restructuring the buyer chooses to do.
2. Fix the accounting policies in the contract
The agreement should say that the earn-out accounts are prepared on the same policies, practices and estimation techniques the business used before the sale, and that where the buyer’s group policies differ, the old ones prevail for earn-out purposes. Without this, the measure can change even though the trading has not.
3. Protect the business as a measurable thing
Negative covenants, meaning things the buyer agrees not to do during the earn-out period: not to move customers or contracts to another group company, not to change the pricing structure without consent, not to transfer key staff out, not to cease or fundamentally change the business, not to merge the accounts with another entity so that the measure can no longer be isolated. A buyer with honest intentions rarely objects to these, because they are describing what they were going to do anyway.
4. Get information rights and a dispute mechanism
You need the right to see monthly management accounts through the period, not just an annual statement at the end. You need a defined window to object, and an independent expert — an accountant, named or appointed by a professional body — whose determination settles the calculation. Without that, disputing a figure means litigation, and the cost of litigation is usually a large fraction of what you are arguing about, which is exactly why the other side can afford to be relaxed.
5. Choose the shape and the window deliberately
Prefer a sliding scale to a cliff. Prefer aggregate measurement across the whole period to annual measurement, because it lets a strong year rescue a weak one. Prefer shorter to longer: the further out the period runs, the less of the result is yours in any meaningful sense. And if there is a threshold, model where your realistic outcome sits relative to it.
The second worked example: the window is worth as much as the target
Two sellers agree an earn-out of up to £400,000 over two years, on an EBITDA target of £500,000 a year. The business then trades identically for both of them:
| Year | Target EBITDA | Actual EBITDA | Percentage of target |
|---|---|---|---|
| Year one | £500,000 | £480,000 | £480,000 ÷ £500,000 = 96% |
| Year two | £500,000 | £540,000 | £540,000 ÷ £500,000 = 108% |
| Two years combined | £1,000,000 | £1,020,000 | £1,020,000 ÷ £1,000,000 = 102% |
Same trading, three structures, three completely different outcomes:
| Structure | Year one pays | Year two pays | Total |
|---|---|---|---|
| A. Annual cliff. £200,000 each year, all or nothing at 100% of target. | 96% — target missed: £0 | 108% — target hit: £200,000 | £200,000 |
| B. Annual slide. £200,000 each year, paid in proportion, capped at 100%. | £200,000 × 96% = £192,000 | Capped: £200,000 | £392,000 |
| C. Aggregate slide. £400,000 measured once across both years, capped at 100%. | 102% of the combined target, capped at the maximum | £400,000 | |
The difference between structure A and structure C is £200,000 — exactly half the entire earn-out — and it was decided by two drafting choices, not by anything anyone did in the business. Structure A punishes a £20,000 shortfall with a £200,000 penalty and gives nothing back for the £40,000 of outperformance that followed. Structure B fixes the cliff. Structure C fixes the cliff and the window together.
If you take one habit from this page, take this one: before agreeing any earn-out, run your realistic case, a 10% worse case and a 10% better case through the exact formula in the contract. It takes twenty minutes. It is the only way to find out whether you have agreed to a sliding scale or to a coin toss.
How earn-outs go wrong, and the order that works
The failures repeat, and they are nearly all visible in the drafting before completion rather than in the conduct afterwards.
- The measure was never defined. “Profit” appears in the agreement with no definition attached. Two years later the seller means one number and the buyer means another, and the contract does not settle it.
- The business was integrated out of existence. The earn-out measures a business that has been merged into a division, whose customers now buy on group contracts and whose costs are shared. There is no clean figure left to measure, and the party holding the money is the party preparing the accounts.
- Costs the seller never agreed to. Group recharges, integration costs, the buyer’s own transaction fees, redundancy from a restructure — all landing on the line the earn-out is calculated from, as the £60,000 charge above did.
- The target was set from the buyer’s optimism. During a sale everyone talks up growth. A target set at the top of the forecast means the seller pays for the buyer’s enthusiasm, because the seller is the one who loses money when it turns out to be enthusiasm.
- The cliff. Missing by a rounding error and receiving nothing, as in structure A above.
- No visibility. The seller has no right to management accounts, so the first time they see the figure is when they are told what it is.
- No dispute route. A disagreement can only be resolved by litigation, which costs a meaningful share of the sum in dispute, so the seller settles for less than the contract entitles them to.
- Employment and earn-out tangled together. The earn-out lapses if the seller leaves, and the buyer decides whether the seller stays.
- The buyer could not pay. An earn-out is an unsecured claim on a company that may have borrowed to buy you. Ask what happens if the buyer is sold on, restructured or fails, and whether any security, parent company covenant or escrow sits behind the obligation.
- Nobody modelled it. Underneath every other failure. The structure was agreed in a meeting and the arithmetic was done afterwards, when it was too late for the arithmetic to change anything.
The order that works
- Value the business on your own figures first. A defensible range, worked privately, before anyone puts a structure in front of you — the free business valuation calculator does this in your browser. An earn-out can only be judged as a share of a number you can already defend.
- Separate every offer into certain and conditional. Cash at completion, fixed deferred, and earn-out. Compare offers on the certain part first, then treat the conditional part as a separate question.
- Fix the measure before the number. Argue about revenue versus gross profit versus a tightly defined EBITDA before you argue about the size of the target. The measure is worth more than the target.
- Model the formula at three outcomes. Plan, 10% below, 10% above — through the exact wording proposed, including threshold and cap.
- Ask what would have to go wrong for this to pay nothing, and then ask how much of that list is inside the buyer’s control rather than yours.
- Then instruct advisers. A solicitor to draft the earn-out schedule, the accounting policies and the protections, and an accountant on the numbers and on how the structure is treated — including its tax treatment, which is a real consideration and belongs with your own accountant, not with a web page.
Reduced to one line: an earn-out is the price of an argument neither side can win with evidence. That is a reasonable thing to agree to. It is only unreasonable when the side that controls the outcome is not the side carrying the risk — and the arithmetic on this page is how you find out which one you have been offered, while there is still time to change it.
Arithmetic and general information only — not financial, tax, legal or investment advice. Your own figures, and your own accountant, decide what any of this means for you.
Do the quick version free
free business valuation calculator — work out the headline value first, because an earn-out is only worth arguing about as a share of a number you can already defend. It runs in your browser and nothing you type is sent anywhere.
The number after debt and cash
The Business Valuation Workbook ($39) goes further than the free calculator above it: five add-back categories rather than owner pay alone, and the step from enterprise value through debt and cash to the equity figure that would reach you. Two linked Excel sheets, run on your own machine. Every row of it is shown on the page before you buy.
See inside the toolkit first
The Business Sale Readiness Toolkit sample shows every sheet, every row of the inputs sheet and the actual Excel formulas — no email, no account. If the sample is not worth your afternoon, the full workbook will not be either.
If you want the readiness assessment too
The Business Sale Readiness & Valuation Toolkit ($499) is a full valuation model with a sensitivity table across multiples, a weighted readiness assessment across the ten areas a buyer examines, a preparation checklist and an adviser prep summary. One-time purchase, instant download.
Frequently asked questions
What is an earn-out when selling a business?
An earn-out is part of the sale price that a buyer pays only if the business meets agreed targets over an agreed period after completion. It differs from deferred consideration, which is a fixed sum the buyer owes whatever happens and only pays later. With an earn-out the amount is not agreed at all, only the formula that will produce it, so it can end up being nothing.
Why do buyers ask for an earn-out?
Because they are paying today for profits that arrive tomorrow, and some of what decides those profits cannot be verified in advance. Owner dependence, recent growth that may not repeat, customer concentration and an unsigned pipeline are all things the seller believes in and the buyer cannot check. An earn-out settles that argument by saying the buyer will pay the seller price if the seller is proved right.
Should an earn-out be based on revenue or profit?
Revenue favours the seller because it is hard for the buyer to influence and easy to verify from the sales ledger. EBITDA favours the buyer because every cost decision the new owner makes, including group recharges and investment, reduces it and therefore reduces the payment. Gross profit sits in the middle, because it stops revenue being bought with discounts but stays above the overheads the new owner controls.
How do you protect an earn-out as a seller?
Move the measure as far up the profit and loss account as the buyer will accept, define in writing exactly which costs may and may not be charged against it, fix the accounting policies as the ones the business used before the sale, and add covenants stopping the buyer moving customers, staff or contracts out of the business being measured. Then add a right to monthly management accounts and an independent expert to settle any dispute over the calculation.
How long does an earn-out usually last?
There is no standard length and it is a negotiable term like any other. The relevant trade-off is that a longer period gives the business more chance to hit its targets but leaves more of the result determined by the new owner decisions rather than the seller own work. Whether each year is measured separately or the whole period is added together often matters as much as the length itself.