Two agencies can bill the same, employ the same number of people and report the same profit, and sell for prices that differ by half. The difference is almost never the quality of the work. It is the answer to two questions a buyer asks in the first meeting: what happens if your largest client leaves, and what happens if you do.
Most agency owners can answer neither with evidence, which is why so many processes stall at the point where the buyer starts modelling. This article sets out how those two risks are actually priced, why they are usually the same risk rather than two, and what an owner can do about them in the year before going to market.
Published ranges in this sector vary more by revenue model than by size, which is unusual and important.
Project-led generalist shops are commonly discussed in the low single digits of EBITDA. Agencies with a mixed book sit in the middle. Retainer-heavy agencies — those with roughly 60% or more of income recurring — are quoted several turns higher, and specialists in defined verticals such as B2B software, healthcare or performance marketing higher still. The spread between the bottom and the top of that range is wide enough that revenue model, not scale, is usually the dominant variable in what an agency is worth.
The buyer universe supports those numbers. Ciesco, which tracks transactions across technology, media and marketing, reported that private equity accounted for 54% of transactions in 2025, driven by platform-led buy-and-build strategies and portfolio bolt-ons, with disclosed value excluding mega-deals rising 61.8% year on year to $91.1bn. The same year saw large-scale strategic consolidation at the top of the market, including Omnicom's $15.9bn acquisition of IPG. Ciesco expects 2026 to be characterised by selective momentum rather than broad-based acceleration.
Read that as an owner. More than half the buyers are financial, buying to build platforms, which means they underwrite on repeatable earnings and on how easily an agency bolts on. Selective momentum means they can afford to be fussy. At Conclave Partners we treat the published multiple as a starting point and spend the preparation effort on the two risks that decide where inside the range an agency lands.
Agencies are unusually exposed to concentration because the work is discretionary, the contracts are short and the client can change supplier without changing anything operational.
Buyers apply rough conventions. A book where no single client exceeds roughly 15% of income is treated as diversified. Above about 25% for the top client, a discount is applied more or less automatically. Above 40%, a meaningful share of buyers decline to proceed at all rather than negotiate a price.
Three refinements matter more than the headline number.
The first is that buyers look at the top five, not just the top one. An agency where no client exceeds 20% but the top five represent 70% of income is concentrated, and presenting only the largest client understates the position — the buyer will build the full table anyway.
The second is that concentration is measured on gross income, not billings. Pass-through media spend inflates revenue without adding value, and an agency that reports billings rather than income will find the picture recalculated on the buyer's basis, usually to its disadvantage.
The third is what sits behind the relationship. A three-year retained contract with a notice period is a different asset from a rolling arrangement with a marketing director who happens to like you. Buyers check contract length, notice, assignment and change-of-control provisions, and they check how long each major client has been with the agency and through how many changes of contact on the client side.
Expect reference calls. Late in diligence buyers commonly ask to speak to two or three major clients, and what those clients say about who they deal with is frequently the single most informative moment of the process.
There is one form of concentration owners rarely think to disclose and buyers always find: sector concentration. An agency whose five largest clients are all in the same industry carries a correlated risk, because a downturn or a regulatory change hits all of them at once. It is worth presenting the sector split alongside the client split, with an explanation, rather than leaving the buyer to notice it unaided.
The second risk is the one owners find hardest to see, because from the inside it looks like commitment rather than concentration.
Valuation practice applies a key person discount where a business depends materially on one individual — conventionally in the range of 5% to 25%, and considerably higher in owner-dependent small companies, where reductions of 30% or more are routinely cited. In agencies the discount tends to sit at the harsher end, because founders typically hold several critical functions at once.
Buyers test four specific dependencies. Who wins the work — if the founder pitches everything, new business stops when they leave. Who holds the client relationship — if the client's marketing director calls the founder directly, the relationship has not transferred. Who makes the creative or strategic calls — if the founder signs off every campaign, quality is personal rather than institutional. And who hires — if the founder is the reason good people join, the talent pipeline is personal too.
The remedy is not delegation in theory but evidence in the numbers: pitches led by others and won, clients whose day-to-day contact is a director rather than the founder, work approved through a documented process. A year of that, visible in the records, moves the discount more than any assurance given in a meeting.
Here is the point most owners miss. In a typical agency the largest client is also the client the founder personally won, personally services and personally reassures when something goes wrong.
That means the buyer is not modelling two independent risks that might partially offset. They are modelling one event: the founder leaves, and the largest client leaves within a year. Where that overlap exists, buyers stop adjusting the multiple and start restructuring the deal — deferring a larger share of the price, extending the earn-out, tying payment explicitly to retention of named clients, or requiring the founder to stay for a period long enough to transfer the relationships.
Separating those two things is therefore the highest-return preparation available to an agency owner. Moving the largest client onto a director's day-to-day management, and having that survive a full year of the client's own staff turnover, changes the risk from one event into two lesser ones. In our experience at Conclave Partners, an agency that has done this is negotiating about price; an agency that has not is negotiating about structure.
The revenue model is the other lever, and it works on both risks at once.
Retained income is contracted, recurring and typically serviced by a team rather than by the founder, because retainers are operational work with rhythms and deliverables. Project income is won repeatedly, often personally, and disappears without notice. That is why the multiple ranges differ so sharply by revenue mix, and why converting project clients onto retainers is worth more than growing project revenue.
The practical work is unglamorous: identify which project clients have continuing needs, propose a scope that reflects what you already do for them, agree a term and a notice period, and put it in writing. Even a modest shift in the mix over twelve months changes the reported recurring share, and the recurring share is the first number a buyer looks at.
Beyond the two risks, agency diligence concentrates on a short list.
Gross income rather than billings, presented consistently across three years, with pass-through costs identified. Buyers value income, not throughput.
Staff cost as a percentage of gross income, which is the sector's central efficiency measure and reveals immediately whether the agency is profitable because of pricing or because of underpayment.
Client and project profitability, not just agency-level margin. Buyers want to see that the largest accounts are the profitable ones, because concentration in a loss-making client is a compounding problem.
Utilisation and freelancer mix, since heavy freelance reliance flatters headcount efficiency while representing capability the buyer does not own.
Owner remuneration restated to a market salary, along with any personal costs run through the business. Adjustments a buyer can verify are accepted; adjustments supported only by the owner's word cause the buyer to discount the whole earnings figure.
Buyers now also ask how the agency uses generative tools and what that means for the fee base. The question cuts both ways and they will want a straight answer to each side of it: which services are becoming cheaper to deliver, and are those savings being kept as margin or handed to clients as lower fees; and which lines of income are most exposed to a client deciding it can produce the same output internally. An agency that has already moved its pricing away from hours towards outcomes is in a stronger position here than one whose rate card still assumes the old cost base. This is not yet a settled area of valuation practice, but it is a standard diligence question, and having a considered answer is better than improvising one.
Financial buyers dominate by volume, as the Ciesco data shows, and in this sector they are typically building platforms: buying one agency as a base and adding specialists around it. They pay well for retained income and clean reporting, and they are the buyers most sensitive to concentration.
Larger agencies and networks buy for capability, client access or geography. They often pay the highest headline price where the fit is genuine, and they integrate quickly, which suits owners who want out rather than those who want autonomy.
Adjacent buyers — technology firms, consultancies, media owners, systems integrators — appear where the agency has a defensible specialism, particularly in performance marketing, data or a regulated vertical.
Management buy-outs work where the senior team already holds the client relationships, which is the same condition that raises the price in any sale. Establishing which of these routes an agency realistically has, before any materials are written, is where Conclave Partners spends the first weeks of an agency mandate.
Move your largest clients off yourself. Introduce a director as day-to-day lead, and let a full year of contact records, meeting notes and invoices show the transfer. This is the single highest-return action available.
Convert project clients to retainers wherever the work is genuinely continuing, with a written scope, term and notice period.
Rebuild the income table: gross income by client for three years, with pass-through costs stripped out, top-one and top-five concentration calculated on that basis, and client tenure shown.
Get the contracts signed. Verbal arrangements with long-standing clients are worth far less than the same relationship with a document, and assignment provisions should be checked before a buyer checks them.
Distribute new business origination. Have someone other than you lead and win pitches, and record who led what.
Document the delivery process — briefing, approval, quality control — so that output is visibly institutional rather than personal.
Restate owner pay to market and clean up personal costs. Sellers who ask Conclave Partners where to begin are pointed at the client-by-client income table and the pitch origination record, because those two documents quantify precisely the risks the buyer is trying to price.
A prepared agency typically takes six to nine months from launch to completion. Financial diligence is quick; the time goes on client contracts, employment terms for the senior team, and negotiating the deferred element.
Expect a substantial part of the price to be contingent. Where concentration and founder dependence are present, buyers manage them through earn-outs tied to client retention rather than through the multiple, and the definitions in those clauses — which clients, measured how, over what period, and what happens if the buyer changes the team — matter more than the headline figure.
Confidentiality is critical. Staff move easily, competitors recruit, and clients who hear rumours start reviewing their roster. A leak damages exactly what is being sold.
Ranges quoted in the market run from low single digits of EBITDA for project-led generalist shops to substantially higher for retainer-heavy agencies with around 60% or more recurring income, with specialists in defined verticals at the top. Revenue model matters more than size in placing an agency within that spread.
There is no single figure, but the conventions are consistent: no client above roughly 15% of gross income is treated as diversified, above about 25% a discount is applied, and above 40% many buyers decline entirely. Buyers assess the top five as well as the top one, and calculate on gross income rather than billings.
It is a valuation reduction applied where a business depends on one individual, conventionally cited in a 5% to 25% range and considerably higher in owner-dependent small companies. Agencies tend to attract the harsher treatment because founders often hold new business, client relationships, creative approval and recruitment simultaneously.
Yes, and materially. Retained income is contracted, recurring and usually delivered by a team rather than the founder, so it reduces both concentration risk and key person risk at once. That is why the quoted multiple ranges differ so sharply by revenue mix.
Usually, and longer where the founder holds the client relationships. The length is negotiable and is shorter for owners who have already transferred day-to-day contact to their senior team. Expect the earn-out to be tied to retention of named clients.
Roughly six to nine months from launch to completion for a prepared business, with client contracts and the deferred consideration the slowest elements. The preparation that changes the price — transferring client relationships and shifting the mix towards retainers — needs a further twelve months to show in the records.
Ildar Zakirov — Conclave Partners ildar@conclavepartners.com
Sergi Kosiakof — Conclave Partners sergi@conclavepartners.com