AGENCY VALUATION
What really drives your number, how buyers value a marketing agency, and how to build toward the figure you want.
Book a conversation

What is your agency worth?
Enter your earnings to see where you sit on our published bands, then answer five short questions. Each answer moves you within your band and narrows the range. Nothing you type leaves your browser.
Indicative only. This applies our published bands and curve to your earnings and your own answers. It is not a valuation, and it is not an independent valuation. A real valuation reads your accounts, tests every adjustment against evidence, measures the five drivers against your own data and takes a view on the buyers in your market. We are not accountants or tax advisers. Take your own advice before acting on any number, including ours.
How we value a marketing agency
Most valuation advice for agency owners stops at “three to eight times EBITDA”. That range is wide enough to be true and too wide to be useful. It leaves three questions unanswered: which earnings figure, which multiple and why that one, and what actually reaches your bank account, which is not the same number.
This is how we answer all three. It is the method we use on live mandates, and every multiple in it comes from bands we publish. The calculator above applies the same bands and curve to your own figure.
1. Which earnings figure: normalised EBITDA
We value on normalised EBITDA (sometimes called adjusted EBITDA), not on revenue, not on gross profit, and not on the profit in your accounts. Normalising means restating your earnings as a new owner would experience them. We apply five categories, the same way every time.
Owner compensation: salary above or below market, family salaries that are not commercial, excess pension, personal insurances, and distributions booked as salary or bonus
Personal expenses through the business: vehicles, travel and entertainment beyond what a new owner would spend, personal subscriptions
Genuine one-off costs: non-recurring legal, restructuring or redundancy costs, a relocation, a one-time IT project, aborted transaction costs
Non-recurring revenue, as a deduction: one-off project work unlikely to repeat, grant income, gains on disposal, revenue from a contract that has ended
Other: used sparingly. If most of the value lands here, the categorisation is wrong
Normalisation runs in both directions. If the owners pay themselves below a market package, that is a deduction, not an add-back, and earnings come down. We see that as often as we see the reverse.
The tests are applied together. Adding back an owner’s personal spending without also testing their salary against market overstates earnings, and a buyer’s accountant will find it in diligence. Every adjustment carries the year it falls in, a one-line reason and the evidence behind it. Depreciation, interest, tax, forecast growth and post-deal synergies are never adjustments.
→ Read more: What is adjusted EBITDA? Isn’t it really just cooking the books?
2. Which multiple: the bands we publish
Multiples of normalised EBITDA, by the size of those earnings:
Micro, under 500k: 3.0x to 4.5x
Small, 500k to 1.5m: 4.0x to 5.5x
Lower mid-market, 1.5m to 5m: 5.0x to 7.0x
Mid-market, 5m to 20m: 6.5x to 10.0x
Strategic asset, strong recurring revenue, clear growth or a rare capability: add one to three turns on top
The bands overlap on purpose. Read them as hard steps and the arithmetic breaks. An agency earning 1,499,000 at the top of Small, 5.5x, would be worth 8.24m. One earning 1,501,000 at the bottom of lower mid-market, 5.0x, would be worth 7.51m. Earning 2,000 more would cost it 739,500.
So we do not read them as steps. We run one continuous curve through the middle of each band, rising with scale on a log scale, because multiples rise with size logarithmically rather than in a straight line. It sits inside its own band at every point and never steps down as earnings grow. On the curve, 500k reads 4.25x, 850k reads 4.63x and 2m reads 5.49x. The calculator uses the same curve.
Where you land inside the band
Scale gets you to a band. Five drivers decide where in it you sit.
Client concentration: the largest client and the top five, as a share of gross profit, grouped by ultimate advertiser rather than by brand line
Revenue durability: the share of revenue that is contracted or genuinely repeats, against project work that has to be won again
Owner dependence: how much revenue sits on a relationship one of the owners personally holds
Growth: measured on the same period as the earnings
Earnings quality: how large the normalisation adjustments are as a share of the result
Together the drivers move a business up to about 15% either side of the curve, and never outside its band. That is why the calculator gives a range around the most likely figure.
3. What you actually receive
Almost every published multiple describes enterprise value, which is what the business is worth. It is not what lands in your account.
Agency deals are normally done cash free, debt free. The buyer takes the business with no cash and no borrowings in it. You keep the cash, you clear the debt, and you deliver a normal level of working capital. So the bridge runs:
Enterprise value
plus the cash in the business
less borrowings and anything the buyer treats as debt
plus or minus the difference between the working capital you deliver and the normal level
equals the equity value to shareholders
What counts as debt is negotiated, not given. A credit card balance, unpaid pension or superannuation, a tax instalment plan, a director’s loan. Each one is an argument, and in most deals they are worth more than a tenth of a turn on the multiple.
Normal working capital cannot be set from one year end. It needs a twelve-month average, because agency working capital swings with billing cycles, and a year end chosen well or badly can move the completion payment substantially.
How the price is paid matters as much: the mix of upfront payment, deferred consideration and earn-out. → Read more: The deal structure playbook
What good looks like on profit
Before thinking about multiples, the market is blunt about EBITDA margin, measured on gross profit.
Under 15%: buyers ask hard questions, and many pass entirely
15 to 20%: acceptable, with little room to negotiate on the multiple
20 to 25%: strong, and the market’s preferred range
Above 25%: exceptional. Fewer than 2% of the agencies we see sit here consistently
A healthy cost structure, as a share of gross profit, is roughly 55% staff, 20% other operating costs and 25% EBITDA. Staff costs consistently above 65% of gross profit read to a buyer as a business carrying too many people, and it is priced accordingly.
EBITDA is the part most within your control. Every extra unit of normalised EBITDA can add five to ten times that to enterprise value.
What this is not
This is a method, not a valuation. A real valuation reads your accounts, tests every adjustment against evidence, measures the five drivers against your own data, and takes a view on the buyers actually in your market at the time.
Anything you calculate from this page is indicative. Our indicative valuations are not independent valuations, we are not accountants or tax advisers, and nothing here is a formal valuation. Take your own advice on tax and structure before acting on any number, including ours.
We give an indicative range as part of a first conversation, before anything is formalised. Bring your last three years of accounts and we will tell you what we see, including the parts you will not like.
Is revenue or profit used to value an agency?
Primarily profit. Normalised EBITDA is the standard basis for agency valuations, a multiple is applied to normalised profit, not to revenue. Revenue matters as context, but two agencies with identical revenue and different profit margins will attract very different valuations. Some buyers also reference revenue multiples as a cross-check, but EBITDA is the primary basis.
What adjustments are typically made to EBITDA?
The most common adjustments include: owner compensation normalisation (where founders pay themselves below or above market salary); one-off legal or advisory costs; market expansion investments; non-recurring R&D expenditure; and non-operational income. Adjustments can move EBITDA up or down, the goal is to reveal the true underlying profitability of the business, not to inflate it. All adjustments are disclosed and must be defensible to a buyer.
How far in advance should I get a valuation?
Earlier than most founders think. A valuation 12 to 24 months before you intend to sell gives you time to act on the gaps it identifies. Many of the factors that drive multiple expansion, management depth, client concentration, revenue quality, take time to improve. A valuation done six weeks before going to market can only tell you what you have; a valuation done two years out can tell you what you could have.
Does my agency need to be audited to get a valuation?
Not to get an indicative valuation, but a formal audit of at least one year is typically required by buyers as part of due diligence. Agencies that have never been audited can still be valued, but the lack of audited accounts will be a factor in buyer confidence and may affect the achievable multiple. Starting the audit process earlier rather than later is advisable.
What's the difference between enterprise value and what I actually receive?
Enterprise value is the headline number. What you receive in cash at close depends on the deal structure, the mix of upfront payment, deferred consideration, and earn-out, as well as any working capital adjustments and transaction costs. Most agency deals involve some earn-out component, linking part of the consideration to post-completion performance. Understanding deal structure is as important as understanding valuation.
How long does it take to sell a marketing agency?
Most transactions take between 6 and 18 months from initial preparation to completion. The preparation phase, financial clean-up, positioning, and CIM development, is the most important and should ideally begin 12 to 24 months before you want to transact. If you're thinking about it, the right time to start a conversation is now.
