Customer concentration: testing the cash effect of one large account

Show the largest customers by revenue, receivables, contribution, and expected future collections. One customer can be a modest share of annual sales but a large share of next month's cash.

Use a consistent period and customer identity. Related accounts may need to be considered together for management risk assessment, while their individual invoices remain separate in the ledger.

Record contract end dates, notice periods, minimum commitments, payment terms, refund rights, and service costs. Have the responsible commercial and legal owners explain uncertain terms.

Separate signed commitments from expected purchases. A customer's past volume is useful evidence but does not guarantee the same orders next quarter.

Model a late payment, reduced order volume, and a lost renewal separately. Each changes different parts of the business. A delayed receipt may leave revenue intact while a cancellation changes delivery costs and future revenue.

For illustration, losing a $20,000 monthly customer with $8,000 of avoidable delivery costs changes contribution by $12,000 under that simplified basis. It does not automatically reduce all company costs by the same proportion.

Review staffing, supplier commitments, inventory, software, and other costs associated with the account. Determine the timing and contractual limits on reducing them.

Keep remaining obligations in the downside model. Assuming every customer-linked cost disappears immediately can overstate the company's resilience.

Consider deposits, credit terms, diversification, pricing, or capacity changes based on the evidence. Avoid imposing a generic concentration limit without considering the company's economics and contracts.

Record the trigger management will monitor and the action it would take. Update the review when an account renews, payment behavior changes, or a new commitment increases exposure. The result should explain which customer event could affect cash and what the team can do before it occurs.