Find the discounts, service time, rework, urgency and payment terms that make a high-revenue customer less profitable than smaller accounts.
Short answer: Large customers often receive lower prices while creating more reporting, revisions, urgent work, credit exposure and senior attention. Build a 12-month profit view for each major customer using realised revenue minus direct delivery cost, account-specific service cost and the value of scarce hours. Rank both total contribution and contribution per constrained hour before you reprice, redesign or leave the account.
Revenue reports reward size. They do not show how difficult the revenue was to earn. A customer invoiced £50,000 can produce less money than one invoiced £20,000 if the larger account receives rebates and consumes disproportionate management time.
Do not assume large customers are bad. Predictable volume can improve purchasing, scheduling and cash generation. The problem is unpriced complexity, especially when concessions accumulate one at a time and no one reviews their combined effect.
Build the Account Profit Waterfall
The Account Profit Waterfall starts with invoiced revenue and subtracts five layers until the economic contribution of the relationship is visible.
- Realised revenue: Include: Invoices less rebates, credits and refunds; Common omission: Quoted or list price
- Direct delivery: Include: Materials, subcontracting, fulfilment and commissions; Common omission: Rush premiums and replacements
- Cost to serve: Include: Meetings, reports, support, revisions and procurement work; Common omission: Senior founder time
- Cash burden: Include: Bad debt, customer-specific finance and collection effort; Common omission: Long approval and payment delays
- Scarce-capacity use: Include: Contribution per specialist hour, machine hour or slot; Common omission: Work displaced by urgent requests
Stop before arbitrary company-wide overhead when comparing one more customer decision. Fixed costs still need recovery across the whole portfolio, but allocating rent by revenue can make a large account look worse without revealing a cost you can change.
My position is that customer profitability should be reviewed before every major renewal, not after service has deteriorated. A customer is not “strategic” merely because losing them would make the revenue chart smaller.
Start with realised revenue
Take the last 12 months of invoices and subtract credit notes, refunds, volume rebates and discounts. If the customer receives free units or unbilled service, value them in the delivery layer.
Use a full buying cycle where possible. One quiet quarter may understate annual volume, while an implementation quarter may overstate service cost. Mark one-off setup separately, but do not remove it if every annual renewal creates a new version.
Check scope against billing. Large customers often gain exceptions through informal messages: another report, a different file format, a rush change or attendance at an extra meeting. Record what was delivered even if the business chose not to invoice it.
Trace direct delivery by account
Assign materials, fulfilment, subcontracting, freight, commission and customer-specific software or compliance checks. Use transaction records or a defensible driver, such as units shipped or translator hours.
Do not allocate an average return rate when the account has its own data. A customer ordering fragile custom items may create three times the replacement cost of a standard account. The average hides the operational decision.
Separate normal production from urgency premiums. If the customer’s late approvals repeatedly trigger express freight or contractor overtime, those costs belong to the account even when staff describe them as general delivery pressure.
Put a price on cost to serve
Ask everyone who touches the account to record time for four representative weeks. Include selling, onboarding, meetings, reporting, support, revisions, collections and senior escalation. Multiply hours by a realistic employment or opportunity cost.
Use cost for work that can be replaced by an employee. Use lost contribution when a founder’s scarce hour displaces billable delivery. Keep the bases labelled rather than combining them invisibly.
Look for service requests generated by your own failures. Rework is still a customer cost in the waterfall because the revenue caused it, but the commercial response is internal improvement, not charging the customer for your mistake.
Measure cash and concentration separately
Long payment terms tie up cash. Map the maximum gap between paying suppliers and receiving customer money, then include any customer-specific borrowing cost and collection time. Do not invent a finance charge where the business uses its own cash, but report the amount and days exposed.
Concentration is a risk adjustment, not an accounting expense. A customer producing 35% of contribution can destabilise the business if they leave. A customer producing 35% of revenue but 8% of contribution may already be exerting more influence than its economic value justifies.
Payment, interest, tax and bad-debt treatment varies by jurisdiction and accounts policy. Ask a qualified local accountant how to record a specific item. Use the management view to make the customer decision without changing statutory records casually.
Worked example: HarbourWord Translation
HarbourWord translates product documentation for exporters. Its largest customer generates £48,000 of annual invoiced revenue. A smaller customer generates £18,000.
- Invoiced revenue: Largest customer: £48,000; Smaller customer: £18,000
- Annual rebate: Largest customer: £3,840; Smaller customer: £0
- Direct translator cost: Largest customer: £26,400; Smaller customer: £8,100
- Rush subcontractor premiums: Largest customer: £3,100; Smaller customer: £0
- Account service time at £32 per hour: Largest customer: £7,040; Smaller customer: £1,920
- Rework and replacement files: Largest customer: £1,750; Smaller customer: £300
- Account contribution: Largest customer: £5,870; Smaller customer: £7,680
For the largest customer, contribution is £48,000 minus £3,840, £26,400, £3,100, £7,040 and £1,750, which equals £5,870. Its contribution rate is 12.2% of invoiced revenue.
The smaller customer contributes £18,000 minus £8,100, £1,920 and £300, which equals £7,680. Its contribution rate is 42.7%. Despite generating £30,000 less revenue, it produces £1,810 more contribution.
The service-time line reveals 220 hours for the large account because £7,040 ÷ £32 = 220. The smaller account uses 60 hours. Contribution after all listed service cost is therefore not the right numerator for another hourly comparison because those hours have already been costed. HarbourWord instead examines the cause: custom weekly reporting, late source files and two approval rounds that are absent from the smaller account.
The response is not an immediate termination. At renewal, HarbourWord prices the weekly report, defines one approval round and adds a rush rule triggered by late source files. If the customer refuses and the capacity can be replaced by work resembling the smaller account, leaving becomes economically rational.
Choose the right remedy
Use a condition-specific price before a broad increase. Charge for the extra report, dedicated meeting, short lead time or non-standard format that creates cost. The customer can then remove the condition rather than simply arguing about your margin.
Where your own process causes rework, fix it and measure the next five jobs. Do not make a customer pay indefinitely for an internal error. Where the customer creates variability, set an input deadline and change-control rule.
Some large accounts produce purchasing or scheduling benefits. Quantify them. If a committed monthly volume reduces material cost by £2,000 across other customers, include that portfolio benefit once, with evidence. Do not use “brand value” as an unmeasured balancing figure.
Exit when the account remains below your minimum contribution after a fair redesign, blocks better work or creates cash risk you cannot fund. Plan the timing and fulfil contractual obligations. Specific termination rights and notice requirements need qualified local legal advice.
Review the portfolio, not one account in isolation
Plot customers by total contribution and contribution per scarce unit. High total and high unit contribution are core. High total but low unit contribution need efficiency or repricing. Low total but high unit contribution may be worth growing. Low on both needs a reason to stay.
Avoid replacing one concentrated customer with another identical one. Set a working maximum share of total contribution for any account and test what happens to cash if it leaves. The appropriate limit depends on contract length, replacement lead time and reserve.
Related guides
Complete the review before the next renewal
This week, choose the five largest customers by revenue and build each waterfall for the last 12 months. During the next four representative weeks, record service time and exceptions by account.
Ten working days before each renewal, identify the two conditions creating the most avoidable cost. Present a revised scope and price tied to those conditions. After 30 days, compare contribution and scarce-capacity use. Retain the account only if the revised economics support the portfolio or a separately measured benefit justifies the difference.
Frequently asked questions
Should I allocate all overhead to each customer?
Allocate fixed overhead when testing whether the whole portfolio is viable, but do not let an arbitrary allocation hide customer-specific causes. For a renewal decision, start with realised revenue, direct delivery, cost to serve and cash burden. Then ask whether the remaining contribution supports the shared structure. Rent divided by customer revenue may be simple, but it does not show what would disappear if the account left. Use both views, label them and avoid calling allocated overhead an incremental cost.
Is a low percentage margin always a reason to drop a customer?
No. A lower percentage can still produce substantial total contribution with little scarce capacity and predictable payment. Compare contribution pounds, contribution per constrained unit, cash exposure and portfolio risk. A 15% account requiring almost no management may be better than a 40% account consuming specialist time. The relationship becomes questionable when low percentage combines with low total contribution, displacement of better work or unmanageable terms. Use both measures together. Require a quantified reason rather than applying one margin threshold to every customer.
How do I value time spent supporting one customer?
Track actual hours by activity for at least four representative weeks. Use employment cost for replaceable support and opportunity contribution for founder or specialist time that displaces paid work. Do not use the customer billing rate automatically because it may include overhead and profit. Annualise only after checking seasonality. If time recording changes behaviour temporarily, repeat the sample. Round cautiously. The aim is a decision-grade estimate, not minute-perfect surveillance. Keep internal failures separate from customer-created requests so the remedy is fair.
What if the customer is important for credibility?
Define what that credibility changes. Does the customer permit a named case study, introduce suitable buyers or satisfy a procurement requirement? Measure the resulting qualified opportunities and contribution over a fixed period. Do not assume a logo creates value, and never display it without permission. Credibility can justify accepting less profit for a limited engagement, but it cannot make negative cash contribution safe. Set an end date and normal renewal terms before the first exception becomes permanent. Review the result after 90 days.
Can I charge more for late information and rush work?
Yes, when the rule is clear, agreed and reflects genuine added cost or capacity disruption. Define the customer input deadline, normal lead time, what counts as a change and the price or revised delivery date that follows. Do not impose a surprise charge after accepting the request. Contract and consumer rules vary by country and buyer type, so obtain qualified local advice on enforceability and fairness. You can also decline the rush rather than charging for risk you cannot manage.
How often should I review customer profitability?
Review major accounts quarterly and before every renewal. Review smaller accounts when scope, service time, payment behaviour or direct cost changes materially. A rolling 12-month view smooths one unusual period, while recent four-week time tracking captures current behaviour. Do not wait for annual accounts if an account is consuming urgent capacity now. Keep the calculation consistent so movement reflects the relationship rather than a new allocation rule. Keep the source data. Investigate changes above a threshold you set from your own margin tolerance.
What should I do if my largest customer refuses new terms?
Compare three numbers before deciding: contribution under current terms, contribution under the proposed terms and contribution expected from the best replacement use of capacity. Include the time and cash needed to replace revenue. You may phase changes, reduce optional service or accept current terms for a defined notice period. Do not concede indefinitely because the invoice value feels frightening to lose. Check contractual notice and delivery obligations with qualified local advice, then choose the option that protects cash and viable contribution.
Comments (0)