Nobody sits down and decides to build a business where one customer can end it with a phone call. It happens the way most bad structural decisions happen — through a sequence of individually sensible yeses.
The client asks for another project, and you say yes because turning down good work feels absurd. They ask you to pick up something adjacent, and you say yes because you need the utilisation. They ask for a retainer, and you say yes because recurring revenue is what everyone tells you to build. Eighteen months later you look at the sales ledger and one logo accounts for more than half of it.
The number that should make you uncomfortable
Put figures on it, because the abstract version never lands. An agency turning over £420,000 with one client at £250,000 is at 60% concentration. Fixed costs — salaries, premises, software, the bookkeeper — run at £26,000 a month, so £312,000 a year.
Lose that one client and revenue drops to £170,000 against a cost base of £312,000. The business is not slightly less profitable. It is losing roughly £12,000 a month, and it will keep doing so for however long it takes to either rebuild the pipeline or make the team redundant.
Now add the notice period. Most agency retainers run on 30 or 90 days. Ninety days sounds generous until you count backwards: it is one quarter to replace 60% of your income, in a market where a new client of that size typically takes two or three quarters to win, sign and start.
A client at 60% of revenue is not a customer relationship. It is an unsecured, uncommitted credit facility that the other side can withdraw with 90 days' notice.
It changes how you behave, long before it goes wrong
The financial exposure is the obvious half. The behavioural half does more damage and shows up sooner.
You stop pricing properly. A scope creep you would push back on with a £3,000-a-month client gets absorbed with the £20,000-a-month one, because the downside of an awkward conversation feels enormous. Over a year that absorbed work is often worth more than the margin on the account.
You stop saying no to bad work, and to bad behaviour. Late payment gets tolerated. Friday-afternoon requests get accepted. Junior staff learn that this client's requests override everything, which is a culture decision you never consciously made.
And you stop selling. Business development is the first thing to be squeezed when the big account is busy, which is precisely the mechanism that deepens the dependency. The account grows because you have no pipeline; you have no pipeline because the account grows.
The other structural detail people miss: the relationship usually rests on one person at their end. When the marketing director who hired you moves on, their successor arrives with their own agency relationships and a mandate to review spend. Nothing you did wrong. The account still goes.
What it does to what the business is worth
This is the part founders discover late, usually in a first conversation with a buyer or a lender.
Concentration is one of the first things a buyer tests, and a business with 60% of revenue in one uncontracted client is not valued on the same multiple as one with its largest client at 15%. Buyers either discount the price, load the deal with earn-out conditional on that client staying, or walk. A lender assessing a facility asks the same question for the same reason, and it feeds directly into what a lender actually asks for before approving a loan.
So the concentration is not just a risk you carry. It is a discount you have already applied to your own exit, years before you plan to take one.
Getting out of it without burning the account
The instinct — fire the big client and rebuild clean — is almost always wrong. It converts a manageable structural problem into an immediate insolvency problem. The route out is slower and duller.
Set a ceiling and measure against it monthly. A practical rule for a small agency is that no client should exceed 25% of revenue, and any client above 35% triggers a specific plan. The point of the number is that it forces the conversation while there is still time to act.
Grow the denominator rather than shrinking the numerator. Winning £100,000 of new work takes the same account from 60% to 48% without a single difficult conversation, and it is the only version of this that improves the business rather than just de-risking it.
Ring-fence sales capacity that the big account cannot touch. One day a week, in the diary, non-negotiable, even in the weeks the client is loudest. If the account is genuinely so demanding that this is impossible, the account is under-priced and you have just proved it.
Contract for the exposure. Push the notice period out to six months at the next renewal, and price the certainty rather than apologising for asking. Get a minimum committed spend written in rather than relying on an expectation.
Then build the buffer that turns a lost account into a bad year instead of a closure. Cash reserve equal to three months of fixed costs is the working target — for the agency above, roughly £78,000. That is a lot, and it is exactly why the reserve has to be built while the big client is still paying, not started once the notice arrives. Knowing the difference between profit and cash is what makes that reserve real rather than theoretical.
The uncomfortable truth is that the moment to fix client concentration is when the relationship is at its best — when they are delighted, paying on time, and expanding the brief. That is the same moment when doing anything about it feels least necessary, which is why so few businesses do.
Common questions
What percentage of revenue from one client is too much?
There is no legal or accounting threshold, but a widely used working rule for small agencies and consultancies is that no single client should exceed about 25% of revenue, with anything above 35% treated as a situation requiring an active plan. The reasoning is simple arithmetic rather than superstition: below a quarter, losing the account is a bad year you can trade through by cutting discretionary costs, while above a third it usually forces redundancies. Contract terms matter too. A client at 30% on a twelve-month committed contract is a smaller risk than one at 20% rolling monthly.
Should I turn down work from my biggest client?
Rarely, and almost never as a first move. Refusing revenue does not reduce the risk you already carry, it just makes the business smaller and less able to absorb a loss. The better response is to accept the work and use the margin to fund the thing that actually fixes concentration, which is business development capacity aimed at everyone else. The exception is work that would require hiring specifically for that client, because that raises your fixed cost base in a way that is directly dependent on one account continuing.
How does client concentration affect selling my business?
It affects both the price and the structure of the deal. A buyer assessing a business where one client represents most of the revenue is really assessing whether that relationship survives a change of ownership, and typically responds by paying a lower multiple, shifting a large part of the price into an earn-out contingent on that client staying, or requiring the founder to remain for a transition period. Because buyers usually look at two or three years of trading history, reducing concentration is something to start well before you plan to sell rather than in the year you go to market.
Does a long contract fix the problem?
It improves it substantially without removing it. A twelve-month committed contract with a six-month notice period converts an immediate cliff into a manageable planning horizon, which is usually enough time to replace a large account. What it does not do is protect you against the client suffering their own financial trouble, restructuring, or being acquired, and a contract is only worth what the counterparty can actually pay. Treat improved terms as buying time to diversify rather than as a substitute for diversifying, and keep watching the concentration percentage regardless.
