Customer concentration measures how dependent a defined sales total is on a small number of buyers. Finish with top-one and top-three shares plus a separate largest-account loss bridge that distinguishes revenue from contribution.
Updated: 29 October 2026.
What does customer concentration tell you?
Customer concentration describes how much of a defined sales total comes from a small number of buyers. This exercise produces two separate outputs: the largest-one and largest-three revenue shares, followed by a conditional loss-of-business bridge. A historical share describes dependence; it does not estimate the probability that an account will leave.
BDC includes major-customer dependence among the matters to investigate during commercial due diligence. We use that question as context, not a universal numerical danger threshold. A fifty-percent sales share is not a fifty-percent departure probability. Nor does a large account automatically justify abandoning a commercially sound relationship.
Every customer code and amount below is fictional training material. It is not an Anar assignment, an actual customer ledger or a measured business result. Arithmetic and saved result files were checked programmatically; no CRM, banking or desktop user trial was performed. The business risk guide assigns responses to events. Our narrower job measures a supplied customer ledger and a specified loss scenario.
Reconcile the period before ranking buyers
Use one completed month, one currency and one consistent revenue basis. In this exercise, figures exclude tax, discounts are already included in the initial sales amounts, and returns are deducted once against the correct buyer. Do not subtract a return twice because it appears in both the sales export and a separate credit report.
Keep a stable customer identifier. Two spellings of one account should not create artificial diversification, but similar names alone do not authorize combining unrelated buyers. Our supplied rule says A, B, C and D are separate customers with no shared group. Real related-account grouping needs evidence rather than a guess about ownership.
| Customer | Initial sales | Returns | Net sales |
|---|---|---|---|
| A | 5,500 | 500 | 5,000 |
| B | 3,200 | 200 | 3,000 |
| C | 1,500 | 0 | 1,500 |
| D | 500 | 0 | 500 |
| Total | 10,700 | 700 | 10,000 |
Check 10,700 minus 700 equals 10,000, then independently add the four net-sales rows. Both routes must agree. A transaction repeated across product lines or payment allocations should not silently become additional sales. Preserve source versions and the reporting cutoff so a later reviewer can reconstruct the population.
The customer lifetime value guide estimates value across a relationship using purchase frequency and lifespan. Our denominator is one frozen period across the customer portfolio. Do not mix lifetime values with monthly revenue to create an impressive but meaningless concentration ratio.
Calculate top-one and top-three shares explicitly
Rank net sales from largest to smallest: A at 5,000, B at 3,000, C at 1,500 and D at 500. Divide each amount by the complete 10,000 net-sales total. Their shares are fifty, thirty, fifteen and five percent. Adding the unrounded shares gives one hundred percent, a useful reconciliation check.
Top1 = 5000/10000 = 50% Top3 = (5000+3000+1500)/10000 = 95% Outside the largest three = 500/10000 = 5%
Top-three means the three largest customers by the chosen net-sales amount, not the latest three orders or the buyers with the most transactions. A customer with many tiny orders need not rank above one with a single substantial contract. Label the basis in the table heading.
The result identifies an evidence question, not a mandatory commercial action. You might review the largest account’s renewal timing, responsible relationship owner or order visibility. This worksheet does not send a message or tell you to reject a valuable buyer. Reducing a share by adding unprofitable business is not automatically an improvement.
Keep money amounts beside percentages. A falling share can result from growth elsewhere while the account’s sales remain unchanged; it can also result from losing that account’s revenue. The percentage alone does not distinguish those stories. Preserve the underlying period totals before comparing future versions.
Separate revenue exposure from contribution exposure
Now create a separate hypothetical future month. Assume the same sales and cost conditions repeat, except that A places no new business. This is not a prediction and does not erase a completed historical sale. It is also not a late-payment scenario: delayed collection and absence of future demand require different models.
| Customer | Scenario sales | Cash costs avoided if that work disappears | Contribution |
|---|---|---|---|
| A | 5,000 | 3,000 | 2,000 |
| B | 3,000 | 1,200 | 1,800 |
| C | 1,500 | 600 | 900 |
| D | 500 | 200 | 300 |
| Total | 10,000 | 5,000 | 5,000 |
Contribution here is sales less the supplied incremental cash costs genuinely avoided when that work is absent. A represents half of revenue but only 2,000 divided by 5,000, or forty percent, of contribution. B represents thirty percent of revenue and thirty-six percent of contribution. The two concentration questions are related but not interchangeable.
Do not assume every expense assigned to a customer disappears on departure. Committed payroll or rent may continue. In this exercise the listed work costs really are avoided within the scenario month; continuing expenses are supplied separately. ACCA’s relevant-cost approach supports examining future cash changes, not merely relabeling accounting allocations.
The profit-margin guide distinguishes gross and net measures relative to sales. This customer-specific contribution table is neither a new net-profit statement nor a universal customer profitability calculation. Document what has and has not been deducted before making a decision.
Build the largest-account loss bridge
With every buyer present, contribution is 5,000 and continuing expenses are 3,500. The model leaves 1,500. Without A, revenue is 5,000, remaining work-dependent costs are 2,000 and contribution is 3,000. Continuing expenses stay at 3,500, leaving negative 500. These are model balances, not certified net profit after every possible tax and expense.
All accounts:10000−5000−3500=1500 Without A:5000−2000−3500=−500 Change:−500−1500=−2000 Independent bridge:−5000 lost sales +3000 avoided costs=−2000
Subtracting all 5,000 lost sales from the original 1,500 would give negative 3,500 and ignore the avoided 3,000. The opposite mistake is deleting continuing rent or salaries merely because an account leaves. Keep both the before-and-after calculation and the independent change bridge so another reviewer can catch either error.
Do not insert replacement sales or immediate cost reductions without supporting assumptions. A recovery plan may deserve a separate scenario, but it must state its own demand, costs and horizon. The cash-flow guide handles when payments occur; a negative-five-hundred model balance does not establish today’s bank balance.
A useful conditional conclusion is: under these fictional repeat-month inputs, losing A reduces the modeled balance by two thousand. It does not say A will leave, that the business fails immediately or that a particular financing product solves the exposure. The next action is to investigate a named missing fact.
Handle ties, small portfolios and invalid denominators
If A and B each have five hundred net sales and no other customers exist, both tie for largest. Top-one is fifty percent. Top-three, using all two available customers, is one hundred percent. Do not invent a third customer or hide the tied account behind an arbitrary sort order. Evaluate each tied account’s loss separately if its avoided costs differ.
A zero-net-sales customer can remain in the source when total net sales are positive; its share is zero. If the entire period total is zero, the percentage is undefined in this simple model. Record not calculated rather than zero concentration. No sales in a period does not prove an absence of commercial risk.
If any customer has negative net sales, stop this nonnegative-share exercise. Do not clip that row to zero or omit it to manufacture clean percentages. Returns relating to earlier periods or adjustments may require a different reporting treatment. Preserve the record and ask the source owner which period and grouping are appropriate.
Unknown customer identities and unassigned returns also prevent a fully reconciled customer ranking. Keep them visible as exceptions. A worksheet that honestly names an unresolved amount is more useful than a precise percentage built by quietly assigning the amount to a convenient buyer.
Save the inputs, result and reversal assumptions
On Windows, search Start for Notepad and create a blank document. Open Calculator separately from Start. Write the period, currency, customer-grouping rule and source names first. Copy the four ledger rows, the sales control and the separate future loss scenario. Fifteen minutes is a planning target, not measured user performance.
Fictional V1:one month; A/B/C/D separate Initial10700−returns700=net10000 Shares A50%,B30%,C15%,D5%;Top1=50%;Top3=95% Contribution5000; continuing3500; model balance1500 Without A:sales5000−workcost2000−continuing3500=−500 Change−2000; independent bridge−5000+3000=−2000 Tie500/500:Top1=50%;Top3=100%; retain both names Zero total or negative customer net sales:STOP percentage model Evidence gaps, owner, next review:[...]
Choose File → Save As, select Documents and save customer-exposure-V1.txt. Use another name if it already exists. Close Notepad, open Documents in File Explorer and reopen the saved file. Locate the ten-thousand control, fifty/ninety-five shares and negative-two-thousand bridge. Save tie and zero tests separately as customer-exposure-V2.txt without overwriting the base record.
Use the decision log to preserve one evidence-based next action: verify account grouping, confirm avoidable costs or review the relationship plan. Retain the hypothetical label and the assumption that all other conditions repeat. No real customer departure, intervention or business improvement is claimed by completing this exercise.
Frequently asked questions
Does a fifty-percent share mean a fifty-percent chance of departure?
No. It describes the account’s share of the selected sales total, not an event probability.
Should all lost revenue be subtracted from profit?
Not without accounting for genuinely avoided costs; expenses that continue must remain in the scenario.
How does top-three work with only two customers?
Use both available customers without inventing a third; their combined share is one hundred percent here.
What if total net sales are zero?
The percentage is not calculated in this model. Do not divide by zero or describe the result as risk free.
