Girişimcilikİş

Client Concentration Risk: When One Big Client Becomes a Liability, and How to Diversify Safely

A self-employed professional at a sunny desk arranging cards into balanced groups, weighing client concentration risk

TL;DR. Public companies treat a single customer worth 10% or more of revenue as a fact investors must be told about (IFRS 8, paragraph 34; FASB ASC 280-10-50-42). Eurostat calls a client “dominant” when it provides at least 75% of a self-employed person’s income, and in its 2017 survey 17.8% of EU self-employed people who had clients worked for one client or a dominant one (CEOtudent calculation from 4.6 million of 25.8 million). Large firms live with heavy concentration too: Apple accounted for about 91% of Cirrus Logic’s net sales in fiscal 2026. The danger for a solo practice is not the big client itself but what happens in the months after it leaves. In a worked example with illustrative assumptions, a practice with a 70% client and costs at 70% of revenue falls $4,000 a month short of its costs the day that client goes. The CEO move: measure top-client share, HHI and the effective number of clients every quarter, and set a ceiling before the market sets one. The student move: read twelve months of your own revenue data, then run one small client-acquisition experiment a month and log what converts.

This article sits in the freelance and client business cluster. It assumes the cash buffer from the runway math and the account structure from the variable income budget, and it treats retainers, covered in the retainer playbook, as one of the main tools for diversifying safely.

This article is general information, not financial, legal or tax advice.

Why does one big client count as a risk at all?

A large client is usually good news when it arrives. The risk is structural: the more of your income one decision-maker controls, the more one budget review or reorganisation can change your year.

Public markets have formalised this. Under IFRS 8, “If revenues from transactions with a single external customer amount to 10 per cent or more of an entity’s revenues”, the company must disclose that fact and the revenue involved. US GAAP has an equivalent rule. In a 2025 comment letter, SEC staff asked one company to disclose “the amount of revenue for each customer that accounts for 10 percent or more of the total” to comply with FASB ASC 280-10-50-42.

Research on listed suppliers shows why investors care. Dhaliwal, Judd, Serfling and Shaikh (2016) report a positive association between customer concentration and a supplier’s cost of equity, and the relation “is more pronounced for suppliers that are more likely to lose major customers” (abstract only). Markets charge more for income that depends on a few buyers. A solo practice pays the same premium in a different currency: lost bargaining power and months of income at risk.

Which thresholds do institutions actually use?

There is no official rule for how concentrated a freelancer’s income should be, but official thresholds for neighbouring questions give a usable scale.

Threshold Figure as published What it measures Source
Major customer disclosure A single external customer at 10 per cent or more of revenue When a listed company must disclose reliance on one customer IFRS 8, paragraph 34 (EU Regulation 2023/1803); FASB ASC 280-10-50-42, as cited by SEC staff, 2025
Highly concentrated market HHI greater than 1,800; a change of more than 100 points is a significant increase Concentration of a market, for merger review US Department of Justice and FTC, Merger Guidelines, 2023
Share presumption A merger creating a firm with a share over thirty percent, plus an HHI increase of more than 100 A second structural test for mergers Same guidelines
Dominant client A client that provided at least 75% of the self-employment income in the last 12 months Economic dependence of the self-employed Eurostat, EU-LFS 2017 ad hoc module metadata
Notice period in practice Customer agreements “generally terminable by our customer upon 30 to 90 days’ notice” How fast a large services relationship can end Innodata, Form 10-K for 2025

Sources: EU Regulation 2023/1803 (IFRS 8); Hallador Energy correspondence with SEC staff (2025); DOJ and FTC (2023); Eurostat (2017 module metadata); Innodata (2026 filing).

Two reading notes. The antitrust thresholds measure markets, not personal income; they are borrowed only because the Herfindahl-Hirschman Index (HHI) is the standard concentration measure. And the “no client above 30%” rule of thumb that circulates in freelance advice has no primary source that CEOtudent could find. The thirty percent figure in the Merger Guidelines is a market-share test for mergers, not a finding about freelance income. Treat any ceiling, including the ones in this article, as a policy you choose rather than a research result.

How concentrated are self-employed people in practice?

The best official dataset on this question is the 2017 EU Labour Force Survey ad hoc module on self-employment, which asked self-employed people how many clients they had and whether one was dominant. It is the most recent edition of that module in Eurostat’s database, so the figures describe 2017, not today.

Area Self-employed with at least one client (thousands) One client only Several clients, one dominant One client or a dominant client
EU-27 25,755.8 9.2% 8.6% 17.8%
Germany 3,322.3 8.7% 8.2% 16.9%
France 2,799.1 5.8% 13.3% 19.0%
Spain 2,850.1 6.6% 6.3% 12.9%
Italy 4,975.8 13.7% 3.1% 16.8%
Netherlands 1,256.7 6.0% 11.5% 17.5%
Poland 2,659.5 9.7% 9.9% 19.6%

Source: CEOtudent analysis of Eurostat dataset lfso_17seclnt (2017, thousand persons). Shares exclude people reporting no clients and non-responses. “Dominant” means at least 75% of self-employment income.

Roughly one in six EU self-employed people with clients sat at or above the 75% line. The dataset includes self-employed people with employees, so the share among solo freelancers may differ. A companion dataset shows the downside from the inside. Among 19.7 million EU own-account workers (self-employed without employees), the most common main difficulty, after “no difficulties” (23.7%), was “periods of having no customer, no assignments or project to work on” (13.8%), followed by high administrative burden (10.7%) and “delayed payments or non-payments” (10.0%), according to CEOtudent calculations from Eurostat dataset lfso_17sediff. Eurostat also identifies a narrower group, the economically dependent self-employed: people without employees who worked for one or a dominant client that “decides about his/her working hours”. In 2017 that group numbered 806,000, or 4.1% of own-account workers.

What does concentration look like in companies that disclose it?

Some of the most concentrated revenue lines in public filings belong to large, profitable companies.

Company and filing Largest customer share, as printed Earlier years Receivables concentration How the filing describes the risk
Cirrus Logic, 10-K for fiscal 2026 Apple, about 91% of net sales 89% (FY2025), 87% (FY2024) Foxconn 39% and Luxshare 22% of gross trade receivables Loss of a key customer “could significantly reduce our sales and our profitability”
Skyworks Solutions, 10-K for fiscal 2026 Apple, 61% of net revenue 67% (FY2025), 69% (FY2024) Three largest balances, 75% of gross receivables Lists “the gain or loss of significant customers” among factors that move results
Innodata, 10-K for 2025 One customer, about 58% of total revenues 48% (2024) 63% of receivables ($29.2 million) due from one customer Contracts “generally do not obligate customers to purchase services in future periods”

Sources: Cirrus Logic 10-K (filed 21 May 2026); Skyworks Solutions 10-K (filed 10 November 2026); Innodata 10-K (filed 26 February 2026).

Three patterns are worth carrying into a solo practice. The share can drift up while the business grows: Innodata’s revenue rose 47.6% from 2024 to 2025 (CEOtudent calculation from $170.5 million to $251.7 million), and its largest customer’s share rose by 10 points at the same time. Concentration in revenue usually means concentration in receivables, so the same client also holds most of your unpaid invoices. And reducing a share takes time: Skyworks moved from 69% in fiscal 2024 to 61% in fiscal 2026. For a services business, the Innodata filing adds the most relevant detail, a 30 to 90 day notice period. That is the window in which a lost anchor has to be replaced.

Is concentration always bad?

No, and the research says so directly. Patatoukas (2012), using a comprehensive sample of supply chain relationships, documents “a positive contemporaneous association” between customer-base concentration and accounting returns, and finds that increases in concentration predict efficiency gains such as lower operating expenses per dollar of sales (abstract only). One deep relationship means less selling and more repeat work, which is why solo operators drift into it.

The fair reading of the evidence is that concentration trades efficiency for fragility. The listed-company research measures firms with many staff, cash reserves and legal teams. A one-person practice gets the efficiency side of the trade in full but has far less capacity to absorb the fragility side. The case for a focused single offer, made in the one-product business, is about concentrating on one product; this article is about not concentrating on one buyer. The two fit together.

How do you score your own concentration?

The scorecard below uses three numbers, all computed from the last twelve months of revenue by client. First, top-client share: the largest client’s revenue divided by total revenue. Second, HHI: square each client’s percentage share and add them up, as the Merger Guidelines define it; the result runs from close to 0 up to 10,000 for a single client. Third, effective number of clients: 10,000 divided by HHI. Five equal clients give an HHI of 2,000 (the guidelines’ own example) and an effective number of exactly 5.

CEOtudent editorial framework: the concentration bands

Band Top-client share What it usually means Default policy
Balanced Under 25% No single exit is a crisis. A top share under 25% also guarantees at least 4 effective clients Maintain. Watch for one client growing faster than the rest
Concentrated 25% to 49% One exit hurts for a quarter or two. Common and manageable Add one mid-sized client before taking more work from the largest
Anchored 50% to 74% The largest client is the business. Pricing power has shifted to the buyer Dilute by addition; build runway equal to the income-at-risk months
Dependent 75% or more At or above Eurostat’s dominant-client line. Economically close to employment without its protections Protect first (notice, buffer), then diversify on a dated plan

The 75% line comes from Eurostat; the 25% and 50% lines are CEOtudent editorial choices, not research findings. Note that by the antitrust yardstick almost every solo practice is “highly concentrated”: anything with fewer than about 5.6 equal clients exceeds an HHI of 1,800. That is expected. For a practice of one, the useful readings are the top-client share and the effective number of clients, with HHI as the arithmetic that connects them.

CEOtudent analysis: the scorecard on five illustrative revenue splits

Revenue split by client (illustrative assumption) Top-client share Top-three share HHI Effective clients Clients at 10% or more Band
A. 70 / 15 / 10 / 5 70% 95% 5,250 1.90 3 Anchored
B. 45 / 25 / 15 / 10 / 5 45% 85% 3,000 3.33 4 Concentrated
C. 30 / 20 / 20 / 15 / 10 / 5 30% 70% 2,050 4.88 5 Concentrated
D. 20 / 18 / 15 / 12 / 10 / 10 / 8 / 7 20% 53% 1,406 7.11 6 Balanced
E. 100 100% 100% 10,000 1.00 1 Dependent

The column “clients at 10% or more” applies the public-company disclosure test to a solo practice. Practice A has four clients but only 1.90 effective clients: three of its four would be “major customers” by the listed-company definition, and one of them is the business.

What happens to the year if the top client leaves?

Concentration only matters through its consequences, so the next table runs the same five practices through one event: the largest client leaves. Every input is an illustrative assumption: $120,000 of annual revenue, $7,000 a month of baseline costs (business costs plus the owner’s salary draw), and six months to replace the lost income.

CEOtudent analysis: the top-client exit test (all inputs are illustrative assumptions)

Practice Annual revenue at risk Monthly revenue after the exit Against $7,000 of monthly costs Income-at-risk months Shortfall over a six-month gap
A. Anchored (70% top) $84,000 $3,000 -$4,000 a month 4.2 $24,000
B. Concentrated (45% top) $54,000 $5,500 -$1,500 a month 2.7 $9,000
C. Concentrated (30% top) $36,000 $7,000 Break-even 1.8 $0
D. Balanced (20% top) $24,000 $8,000 +$1,000 a month 1.2 $0
E. Dependent (100% top) $120,000 $0 -$7,000 a month 6.0 $42,000

Two rules fall out of the arithmetic, and both work with your own numbers.

The danger line is one minus your cost ratio. With costs at $7,000 against $10,000 of monthly revenue, the cost ratio is 70%, so any client above 30% of revenue pushes the practice below its costs the moment it leaves. A leaner practice with costs at 50% of revenue can carry a 50% client before an exit creates a monthly deficit. This is why the right ceiling is personal: it depends on your cost base, not on a universal percentage.

Income-at-risk months equal top-client share times replacement months. A 70% client and a six-month replacement gap put 4.2 months of total revenue at risk. If your runway covers fewer months than that, the concentration is uninsured. The runway math linked above shows how to estimate a realistic replacement time; this test tells you how much of it the largest client exposes.

How do you diversify without destroying the income you have?

The most common mistake is to treat diversification as subtraction: dropping or squeezing the big client to force the share down. That cuts income first and hopes the market fills the gap. The safer path is dilution by addition, where the anchor’s revenue is held while the rest of the practice grows around it.

CEOtudent analysis: how much new revenue dilution requires (anchor revenue held constant)

Top-client share today Target share Total revenue must grow to New revenue from other clients, as % of today’s total On $120,000 today (illustrative assumption) New clients needed at 10% of today’s revenue each
80% 50% 1.60x 60% $72,000 6.0
70% 50% 1.40x 40% $48,000 4.0
70% 40% 1.75x 75% $90,000 7.5
60% 40% 1.50x 50% $60,000 5.0
50% 30% 1.67x 67% $80,000 6.7
45% 25% 1.80x 80% $96,000 8.0

The table explains why diversification feels slow. Taking a 70% client down to 40% by addition alone means adding three quarters of today’s revenue from other clients. For a solo operator, that is rarely possible on hours alone. Three levers make it feasible, and they are best combined.

  1. Raise the price of new work. New clients can be priced above the anchor’s legacy rate, so each new client counts for more of the gap. The repricing playbook covers how to do this without damaging existing relationships.
  2. Narrow the anchor’s scope, not its rate. In an illustrative case, cutting the anchor’s work by a quarter while holding its rate reduces the new revenue needed to reach 40% from 75 points of today’s total to 48.75 points (CEOtudent calculation), and frees hours for the pipeline. Total income still ends above today’s level. Scope creep economics explains why the unbilled extras in a large account are often the first hours to reclaim.
  3. Add income that is not another client. A product, a course or licensing income dilutes the anchor without competing for the same hours. The revenue stack lays out the options and their startup costs.

What is the de-risking sequence, band by band?

The order matters as much as the actions. Protect first, then dilute, then lock in.

CEOtudent editorial framework: the staged de-risking sequence

Band today Stage 1: protect (first 30 days) Stage 2: dilute (next 3 to 12 months) Stage 3: lock in Signal you have moved down a band
Dependent (75% or more) Put the relationship in writing: scope, payment date, notice period. Build a buffer toward the income-at-risk months. Check how your national rules treat an arrangement where one client sets your hours Reserve a fixed block of the week (for example one day) for pipeline work, protected from the client. Aim for one paying second client before anything else Make the second client recurring Top share under 75% for two consecutive quarters
Anchored (50% to 74%) Ask for a notice period that covers your replacement time. Separate the anchor’s invoices from the rest so you see its payment behaviour Route all new capacity to non-anchor clients. Price new work at your current rate, not the anchor’s legacy rate Convert one or two project clients to retainers Top share under 50%; effective clients above 2
Concentrated (25% to 49%) Confirm runway covers top share times replacement months Add one mid-sized client per half-year. Decline expansions from the largest client that would push it back above 50% Review the client list every quarter Top share under 25%
Balanced (under 25%) None needed Grow the clients that pay best, not the ones that ask most Quarterly scorecard; act on drift Stays under 25%

The retainer playbook linked at the top has the conversion ladder for Stage 3. New clients also need a channel: the channel economics comparison sets out what marketplaces and direct work actually cost, which affects where the next client should come from.

Why do payment terms belong in a concentration plan?

A concentrated practice is exposed twice: once to the client leaving, and again to the client paying late, because its invoices are most of your receivables. The listed companies above show the same pattern; Innodata reported 63% of receivables due from one customer.

Rules vary by jurisdiction, but two published rules show what policymakers treat as normal. In the EU, Directive 2011/7/EU, which EUR-Lex lists as in force, provides that business-to-business payment periods should as a general rule be limited to 60 calendar days unless expressly agreed otherwise and not grossly unfair, that public authorities should normally pay within 30 calendar days, that statutory late-payment interest is the reference rate plus at least eight percentage points, and that creditors can claim a fixed EUR 40 for recovery costs. In New York City, the Freelance Isn’t Free Act requires a written contract for work worth $800 or more (including agreements totalling $800 in any 120-day period) and, where the contract sets no payment date, payment within 30 days after the work is completed.

The anchor’s payment date is part of its price: a 60-day term on a client worth 70% of revenue means always carrying more than a month of total income as unpaid invoices. The variable income budget linked at the top shows how to keep that gap out of your personal spending.

The tool: a quarterly concentration review and a decision test

CEOtudent editorial framework: the 20-minute quarterly concentration review

  1. Export twelve months of invoiced revenue by client from your accounting tool or bank records. Use invoiced amounts, and note any client more than 30 days late.
  2. Compute each client’s share, the top-client share, the top-three share, HHI and the effective number of clients.
  3. Place yourself in a band. Compare with the previous quarter: which client grew fastest?
  4. Compute your danger line (one minus your cost ratio) and your income-at-risk months (top share times your realistic replacement months).
  5. Compare income-at-risk months with your runway in months. If runway is shorter, the first action is buffer, not marketing.
  6. Write down the notice period and payment term of every client above 10% of revenue. If you cannot find them in writing, that is the finding.
  7. Pick one acquisition experiment for the next month (see below) and one number to judge it by.

CEOtudent editorial framework: the “more work from the biggest client” decision test. Before accepting an expansion from your largest client, answer six questions. Two or more “yes” answers mean negotiate terms first or decline.

Question Yes means
Would the extra work move the client into a higher band? You are trading resilience for revenue; price it accordingly
Would it take hours you currently spend on pipeline work? Diversification stops; set a protected pipeline block first
Is the notice period shorter than your realistic replacement time? Ask for a longer notice period or a minimum term before expanding
Are payment terms longer than 30 days, or is the client already paying late? Your receivables concentration rises with the work
Is the extra work at the client’s legacy rate rather than your current rate? Expansion locks in an old price at larger volume
Would your income-at-risk months exceed your runway after the expansion? The exposure is uninsured; build buffer first

The CEO and the student in client concentration

The CEO lens is portfolio thinking. A chief executive would not let one customer own most of revenue without pricing it and planning for its exit. For a practice of one, that means to set a top-client ceiling that matches your cost ratio, to hold a buffer sized to income-at-risk months, and to treat every expansion from the largest client as a capital allocation decision rather than a compliment. The antifragile individual makes the broader case: cap the downside first, then keep options open.

The student lens is learning to read your own data and to run small experiments. Many solo operators can name their biggest client but not its share, or whether it rose this year. The quarterly review fixes that. The experiments do the rest: one channel or offer per month, a single metric (conversations started, proposals sent, first invoices issued), and a written note on what you learned. Over a year that produces twelve tested answers to the question “where does the next client come from?”, which is the only thing that makes diversification more than an intention.

What this does not mean

It does not mean a big client is a mistake. The research links concentration with efficiency; the aim is to price and insure the dependence, not to avoid it.

It does not mean there is a universal ceiling. The 25% and 50% bands are editorial choices; your defensible ceiling comes from your cost ratio and runway.

It does not mean antitrust thresholds apply to freelancers. The HHI is used here as arithmetic only.

It does not mean the EU figures describe today. The Eurostat module is from 2017 and covers all self-employed people. It shows that the problem is common, not how common it is in 2026.

And it does not replace legal advice on dependent self-employment, which depends on national rules.

Frequently asked questions

What is client concentration risk?
The risk that a large share of income depends on one client, so one decision by that client can change your year. It is measured by top-client share and, across all clients, by the Herfindahl-Hirschman Index.

What percentage of revenue from one client is too much?
No primary research sets a universal number for freelancers. Listed companies must disclose any customer at 10% or more, and Eurostat calls a client dominant at 75% of income. A practical personal limit is your danger line: one minus your cost ratio. With costs at 70% of revenue, that is 30%.

How is HHI calculated for a freelance business?
Take twelve months of revenue by client, convert each to a percentage share, square each percentage and add the squares. Divide 10,000 by the result to get your effective number of clients. Four clients at 70, 15, 10 and 5 percent give an HHI of 5,250, or 1.90 effective clients.

Should a freelancer drop the biggest client to diversify?
Usually not. Dropping it cuts income before replacements exist. Hold the anchor, route new capacity elsewhere, price new work higher and narrow the anchor’s scope rather than its rate.

How long does it take to reduce dependence on one client?
It depends on how fast new revenue arrives. Moving a 70% client to 40% by addition requires new revenue equal to 75% of today’s total, or 7.5 new clients each worth 10% of today’s revenue. Even a large listed company, Skyworks, needed from fiscal 2024 to fiscal 2026 to move its largest customer from 69% to 61%.

Is working for one client the same as being an employee?
Statistically, Eurostat separates the economically dependent self-employed: people without employees who work for one or a dominant client that decides their working hours. Legally, the answer depends on national rules, which this article does not cover.

Does a retainer with the biggest client reduce risk?
A retainer with the anchor improves predictability but can deepen concentration. Retainers reduce concentration risk when they are signed with several smaller clients, and when the anchor’s retainer includes a notice period that covers your replacement time.

Sources

  1. US Department of Justice and Federal Trade Commission (2023). Merger Guidelines, 18 December 2023. Section 2.1 and footnotes 13-15.
  2. European Commission (2023). Commission Regulation (EU) 2023/1803 of 13 August 2023 adopting certain international accounting standards, Official Journal L 237, 26 September 2023. IFRS 8 Operating Segments, paragraph 34.
  3. Hallador Energy Company (2025). Response letter to the SEC Division of Corporation Finance quoting staff comment on FASB ASC 280-10-50-42, filed on EDGAR 28 August 2025.
  4. Eurostat (2024 update). Self-employed persons by number and importance of clients in the last 12 months (2017), dataset lfso_17seclnt.
  5. Eurostat (2024 update). Self-employed persons by main difficulty as self-employed and professional status (2017), dataset lfso_17sediff.
  6. Eurostat (2018). EU-LFS ad hoc module 2017 on self-employment, reference metadata (lfso_17_esms): definitions of dominant client and economically dependent self-employed.
  7. Cirrus Logic, Inc. (2026). Annual Report on Form 10-K for the fiscal year ended 28 March 2026. SEC EDGAR.
  8. Skyworks Solutions, Inc. (2026). Annual Report on Form 10-K for the fiscal year ended 2 October 2026. SEC EDGAR.
  9. Innodata Inc. (2026). Annual Report on Form 10-K for the fiscal year ended 31 December 2025. SEC EDGAR.
  10. Patatoukas, P. N. (2012). Customer-Base Concentration: Implications for Firm Performance and Capital Markets. The Accounting Review, 87(2), 363-392. (abstract only)
  11. Dhaliwal, D., Judd, J. S., Serfling, M., and Shaikh, S. (2016). Customer concentration risk and the cost of equity capital. Journal of Accounting and Economics, 61(1), 23-48. (abstract only, SSRN working paper version)
  12. European Parliament and Council (2011). Directive 2011/7/EU on combating late payment in commercial transactions, Official Journal L 48, 23 February 2011; status as listed on EUR-Lex.
  13. New York City Department of Consumer and Worker Protection. Freelance Isn’t Free Act: workers’ rights page.

Note on method: the EU client-structure shares and the main-difficulty shares are CEOtudent analyses of Eurostat datasets; the company trend figures are as printed in the filings, with the revenue growth figure calculated by CEOtudent. The concentration bands, the de-risking sequence, the quarterly review and the decision test are a CEOtudent editorial framework. The scorecard, the exit test and the dilution table are CEOtudent analyses built on illustrative assumptions ($120,000 of revenue, $7,000 of monthly costs, a six-month replacement gap and the client splits shown); they are worked examples, not data. CEOtudent ran no survey. Deliberately not used, because no primary source was found: the “never let one client exceed 30%” rule as a research finding, and widely repeated claims about how many clients freelancers typically have. Freelance surveys published by platforms and staffing firms were not used because their sampling could not be checked.


This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.

This post is also available in: Türkçe Français Español Deutsch

Benzer içerikler