The Customer Who Owns Your P&L: Stress-Testing Concentration Risk in 2026

Most SME leaders know roughly what proportion of revenue their largest customer provides. Far fewer know what proportion of contribution margin, cash and productive capacity that same customer accounts for. The difference between those two figures is where concentration risk lives, and in 2026 — with buyer consolidation, longer payment terms and thinner buffers across many supply chains — it deserves a place on the board agenda rather than in the category of things everyone already understands.

The purpose of this article is not to argue that concentration is inherently bad. In many UK SMEs a single anchor account is the reason the business exists, funds its investment and supports a level of service that smaller customers could never pay for. The purpose is to make the exposure visible, quantified and manageable, so that it is a deliberate commercial position rather than an accident of growth.

Concentration is rarely a revenue problem first

The headline percentage is the easy part. A customer representing 35 per cent of turnover sounds manageable until three further questions are asked.

What share of contribution does that customer provide? A large account won on price often delivers a materially lower margin than the rest of the book, and occasionally delivers negative contribution once service demands are properly costed. Conversely, an account that is only 20 per cent of revenue may be 45 per cent of profit.

What share of cash does it represent? If the account pays on 75-day terms while the rest of the ledger pays in 30, its share of the debtor book will exceed its share of sales, and its share of financing cost will exceed both.

And what share of capacity does it absorb? Dedicated tooling, a named account team, priority slots in the production plan, bespoke stock. Those commitments are difficult to redeploy quickly and they are the reason losing an account can feel more violent than the revenue figure implies.

Start with arithmetic, not strategy

A simple worked illustration. Take an SME with £8m of turnover and £900k of operating profit. Its largest customer provides £2.4m — 30 per cent of revenue — at a contribution margin after variable costs of 20 per cent, against 28 per cent across the rest of the book. That is £480k of contribution.

Lose the account and the contribution goes with it. The overhead base, the lease, the salaried team and the finance costs do not. Unless roughly £480k can be taken out of the cost base within a year — which in most SMEs means a painful and slow process — the operating profit does not fall by 30 per cent. It falls by more than half.

The arithmetic is deliberately blunt, because the conversation it prompts is the valuable part. What is the minimum overhead we could carry at two-thirds of current revenue? What could we stop doing, and how quickly? Is any of that fixed cost actually discretionary once notice is served?

Four stress tests worth running this quarter

The sudden-stop test. Model the loss of the largest account from a given date with no notice. How many months of cash does the business hold, allowing for the working capital that unwinds as that account's debtors settle and its creditors fall due? If the answer is fewer than three months, that is the finding — not the revenue percentage.

The slow squeeze. Rather than losing the customer, imagine it demands a 5 per cent price reduction and moves to 90-day terms. On the illustration above, that is a direct hit to contribution plus an extra 45 days of funding on the affected sales. Many SMEs survive the sudden loss in a spreadsheet but are slowly bled by the squeeze, because it never triggers a single decisive conversation.

The covenant test. UK lenders increasingly look at customer concentration when setting facilities and pricing. Ask directly how the bank models it, and check whether any facility agreement contains a material-customer clause. Discovering a covenant test after the fact is an avoidable problem.

The capacity test. If the account disappeared tomorrow, which people, assets and slots would be idle, and for how long? Some can be redeployed within weeks; specialised equipment or a dedicated shift pattern may take a year. The redeployment timeline is the real recovery timeline.

Warning signs in pricing and contracts

Behavioural signals usually precede financial ones. Terms creeping from 30 days to 60 and then to 90. A request for retrospective rebates, or a benchmarking clause that allows the customer to demand the lowest price in the market. Procurement processes that shift decision-making to a central team with no relationship to your operational contacts. A single point of contact on your side who handles everything for the account, so that no one else in the business understands it.

Then read the contract for three things: the notice period, the price review mechanism, and whether there is any commitment at all beyond a rolling monthly arrangement. A contract with no notice period and no indexation clause transfers almost all of the risk to the supplier. An annual uplift linked to a published index — CPI, RPI or an agreed sector measure — is not aggressive; it is simply the minimum defence against input cost inflation.

A contingency plan that fits in a quarter

Diversification is the correct long-term answer and the wrong immediate one. A five-year plan to reduce concentration to 15 per cent is unlikely to survive contact with a business that needs to hit its numbers this year. What works instead is a set of contained, sequential actions.

First, know the number. Write down the cash burn and the minimum viable cost base if the account left. Agree it with the finance lead and, ideally, with the bank. Shared arithmetic is far less alarming than a surprise.

Second, protect the relationship commercially. Fix the notice period, insert an indexation clause at the next renewal, and move to stage payments or deposits if the account is large and bespoke. None of this requires a confrontation; it requires preparation before the renewal meeting rather than during it.

Third, grow the second tier deliberately. Identify three or four existing customers where share of wallet could realistically double without inventing a new product or market. A modest, sustained lift across that tier is usually more achievable than a new-logo campaign and it moves the concentration ratio just as effectively.

Fourth, keep an option open. Preserve some cross-training on the account, resist dedicating the last available engineer to it, and avoid letting one customer's bespoke requirements become the only thing the operation knows how to do.

Fifth, rehearse it. Once a year, spend ninety minutes at a leadership meeting walking through the sudden-stop scenario as though it had happened that morning. Who calls the bank? Who speaks to the staff? Which costs go first? Teams that have rehearsed recover far faster than teams that have only modelled.

The balanced conclusion

Concentration is often the price of a genuinely good commercial relationship, and paying it can be rational. What is not rational is paying it unknowingly. A board that can state its exposure in revenue, contribution, cash and capacity terms — and that has a rehearsed plan for the day the account leaves — is in a materially stronger position than one that has simply decided not to think about it. In 2026, with margins tight and buyers consolidating, that difference in preparation is likely to matter more than the percentage itself.