Customer concentration risk is the exposure that builds when one customer becomes too much of your revenue. The working thresholds: under 15% of revenue is normal, 15% to 30% is a dependency worth managing, and past 30% the customer is effectively a partner whose business decisions are now your risk.
The risk has a special shape: it arrives as good news. The logo that makes investors nod, the contract that doubled your revenue, the account your whole Slack celebrates. Twelve months later, that same account is the reason you cannot sleep, because it has quietly become 40% of your income, and its procurement department just went silent.
Concentration is not a moral failing. Almost every company that survives its first years does it by over-serving a few customers who pay well. The failure is not knowing the number, and not pricing it into your decisions.
Measure it in one minute
Take your monthly recurring revenue by customer and compute each one's share. Then look at three thresholds:
| Share of revenue | What it means |
|---|---|
| Under 15% | Normal. Lose them and it hurts, but it is a bad quarter, not an event. |
| 15% to 30% | A dependency. Their renewal date belongs on your calendar, not just theirs. |
| Over 30% | A partner, whether you chose that or not. Their business decisions are now your risk. |
If one customer is over 30%, run your runway math twice: once as-is, and once assuming they leave with their notice period. The second number is your real risk position. If losing them takes you under six months of runway, act now, not at renewal time.
Why it behaves like debt
A dominant customer collects interest in ways that never appear on an invoice:
Product debt. Their feature requests win every prioritization meeting, because of course they do. Two years in, you have built their internal tool, not your product.
Pricing debt. You cannot raise their price, because the downside of the conversation is existential. So your biggest account is often your worst-priced one.
Payment terms debt. Big companies pay in 60 or 90 days, and you accept it, because what else? Meanwhile payroll runs monthly. Concentration plus slow payment is how profitable companies die of cash flow.
Reducing it without heroics
You do not fix concentration by firing your best customer. You fix it by growing the denominator and hardening the terms:
1. Set a ceiling for new deals. A simple rule like "no new customer above 25% of MRR at signing" forces the pricing or phasing conversation early, when you still have leverage.
2. Contract the risk down. Longer notice periods, annual commitments, or prepayment discounts convert fragile revenue into durable revenue. A 10% discount for annual prepay is often the cheapest insurance you can buy.
3. Point marketing at the clones. The fastest path to customer eleven is customer ten's twin. Write down what makes the big account a great fit, and go find three more of the same shape, smaller.
4. Keep the buffer honest. While concentration is high, hold more cash than the standard advice says. Six months of runway assuming the whale leaves is a very different company than six months assuming they stay.
Watch it monthly, not annually
Concentration drifts. A few expansions on one account, some churn among the small ones, and the number you checked in January is fiction by June. Put it next to your runway as a monthly glance. Plainhub flags it automatically when any customer crosses 30% of MRR, but a spreadsheet column works too, as long as someone actually looks.
Celebrate the big logo. Then measure it, contract it, and clone it, so it stays a customer and never becomes your board.