Measure Customer Concentration Risk Before You Sell

Give Each Revenue Share a Denominator

Customer concentration means how much of a defined revenue total comes from a client or group of clients. Rank accounts over the same reporting period, then calculate each account's share of total revenue and of managed services revenue. Combine related client entities where the same commercial decision could remove them together.

The MSP model's US example has $3,634,615 of annual revenue, including $1,890,000 of managed services revenue. Its 35 managed clients average $4,500 in MRR. That average tells you nothing about the largest account without the individual client list. Very different distributions can produce the same average.

For a hypothetical stress test, put an account at the top of the model's $3,000 to $7,000 client MRR sweet spot. Calculate its annual fee as $7,000 × 12 = $84,000. Assuming no other purchases from that account, divide that by each example revenue total: $84,000 ÷ $3,634,615 is about 2.3%, while $84,000 ÷ $1,890,000 is about 4.4%.

Those percentages answer different questions. If the account also buys hardware and projects, add those purchases to its total-revenue numerator. Keep their costs and margins separate when testing earnings. A managed fee alone cannot describe the whole relationship.

Build the Loss From Costs You Can Remove

The model targets a 50% fully loaded managed services gross margin, after burdened delivery labor and tools. Applying that target to the hypothetical $84,000 account would allocate $42,000 of delivery cost and leave $42,000 of gross profit. Neither allocation proves what would disappear if the account left.

EBITDA is earnings before interest, tax, depreciation, and amortization, after operating costs and a fair-market owner salary. For a client-loss scenario, calculate the reduction in EBITDA as lost revenue minus costs actually avoided. Include any additional transition costs in the period when you expect to incur them.

A technician's available hours may increase after a departure while their salary stays exactly the same. Shared account management and tooling commitments may also remain. Freed capacity can serve another account, but it isn't a saving until a cost is removed or replacement work produces income.

• Have finance identify charges that stop, their cancellation terms, and the effective dates. Distinguish client-specific purchases from shared commitments.
• Have the service lead identify the work that stops and the coverage the remaining clients still need. Show any proposed staffing change separately from hours becoming available.
• Have the owner record retained overhead, transition spending, and replacement revenue assumptions. Give each assumption an owner and a date for checking it.

Run the immediate loss and the later operating plan separately. Otherwise a cost reduction that needs time to implement can appear to fund payroll before it exists.

Recalculate Value at the Lower Earnings

Assume, for this limited stress test, that the hypothetical $84,000 account leaves the US stable example, no costs are avoided during the period, and all other revenue and costs stay unchanged. There is no transition spending or replacement sale in this calculation. Calculate stress-test EBITDA by subtracting the assumed lost revenue from starting EBITDA: $730,276 − $84,000 = $646,276.

This is a retained-cost scenario, not a forecast of the account's actual departure. Once finance establishes avoidable costs, add them back to the $646,276 result. If there are transition costs, subtract them for the relevant period. Don't replace that work with the allocated $42,000 margin.

Then recalculate the multiple. The model's valuation curve uses 4.75 times EBITDA at $500,000 and 6.5 times at $1,000,000, with straight-line interpolation between them. For this hypothetical calculation, use 4.75 + (($646,276 − $500,000) ÷ ($1,000,000 − $500,000)) × (6.5 − 4.75). The calculated multiple is about 5.262, but carry the unrounded result into the next step.

To calculate value in this hypothetical stress test, multiply $646,276 by that unrounded multiple: about $3,400,682. Subtract that result from the model’s $4,057,386 starting value for a calculated difference of about $656,704. Both earnings and the size-dependent multiple have changed.

Those are the model's iTValuations 2025 market assumptions, from a seller's market, with yearly review. They aren't a current buyer offer or sale proceeds. Use the curve to expose sensitivity, then have your adviser test the market assumptions before using them in a sale decision.

Judge the Account Before Changing It

My view: a large profitable account deserves a dependence plan, not an automatic exit. A large low-margin account needs an operating review as well. Revenue share alone cannot tell you which case you have.

Grade the account on margin and fit alongside the rest of your clients each quarter. Count the delivery work that can otherwise disappear into overhead, including owner technical work, account planning, and project presales. Keep a record of costs you allocate to assess account economics and costs you could actually remove after losing it.

Where the environment drives excessive delivery work, scope a corrective project before repricing. Then test whether the revised service can earn its required margin. If you decide the relationship must change, work out how the business will replace weak revenue before assuming it can afford the loss.

Reduce Dependence Without Buying Another Problem

Use the concentration list when choosing the rest of the sales pipeline. Winning suitable, profitable accounts can reduce the existing client's share, provided the new revenue actually arrives and the delivery costs are funded. A larger denominator built from underpriced contracts is an expensive way to improve a percentage.

Keep the existing account's agreement, scope, invoicing, and relationship responsibilities current while you build that pipeline. Name who understands the client's business priorities and who maintains technical standards. These are responsibilities you can assign; they don't guarantee renewal.

Take your largest account through the retained-cost calculation now. The useful output is a dated plan showing the revenue at risk, the costs that would remain, the changes you could make, and the earnings left afterward. That gives you something more useful than a reassuring average client size.