If one client funds a third or more of monthly revenue, you do not only have an account. You have a single point of failure wearing a friendly Slack channel.
Concentration risk is quiet until it is not. A budget cut, a new CMO, or a delayed invoice can move the whole studio from busy to fragile in a quarter. The fix is not paranoia. It is measurement, honest thresholds, and diversification that respects the account that got you here.
What client concentration risk means
Client concentration risk is the vulnerability that comes when too much of your agency’s revenue, margin, or capacity depends on one client — or a very small set of clients. It is not a moral judgment on a good account. It is a structural fact about how the business would behave if that relationship changed.
Concentration shows up in more than one number:
- Revenue share — what percentage of invoiced work one client represents
- Gross profit share — which clients actually carry overhead after delivery cost
- Capacity share — how much of the team’s productive hours one logo consumes
- Cash timing — whether a large client also stretches payment terms
A client can look “healthy” on revenue and still be dangerous on profit if the team routinely absorbs unpaid scope. A mid-size retainer can be riskier than a larger project client if every deadline bends around that one account.
Independent agencies often discover concentration only at renewal. By then, hiring, rent, and tool contracts may already assume the big client stays. Naming the risk early turns a surprise into a plan.
A scenario: the account everyone protects
Imagine a twelve-person studio. One B2B client has been with you for three years. The relationship is warm. Work is steady. Their retainer covers a meaningful share of payroll.
Over time, small concessions compound:
- Rush requests skip the normal queue
- Strategy hours bleed into production without a change order
- You delayed hiring for a specialist role because renewal felt “basically done”
- The founder handles most senior contact; nobody else has full context
Then the client hires a new marketing lead who wants to “reset the roster.” Or finance pauses the retainer for one quarter while they renegotiate internally. Nothing dramatic — just a gap you did not price for.
In that moment, concentration is not an abstract metric. It is whether you can make payroll without firing the people you kept busy for one logo. The team feels the shock. The founder feels the guilt. The client may still be reasonable. The business design was the problem.
This pattern is common enough that it deserves a monthly check, not a annual panic.
Measure it every month
Calculate each client’s share of trailing three-month revenue and of gross profit (revenue minus direct delivery cost you can attribute). Revenue share hides unpaid overtime; profit share tells you who actually carries the studio.
Add capacity if you track time: hours per client as a share of billable hours. Two clients at 15% revenue might be fine; one at 15% revenue and 40% of senior hours is a different story.
Write the top three shares on the same page as cash. If you only discover concentration at renewal panic, the measurement arrived too late.
A simple monthly ritual:
- Export invoiced revenue by client for the last three months
- Note gross profit per client if your finance system supports it
- Rank clients by share and flag anyone crossing your thresholds
- Record one behavioural signal (queue-jumping, scope drift, payment drift)
Fifteen minutes is enough if the data is already in one place.
Threshold bands (starting points)
Useful starting rules for an independent agency:
| Band | Revenue share (trailing 3 months) | What it usually means |
|---|---|---|
| Green | No client above ~20% | Healthy spread; still watch profit and capacity |
| Amber | One client between ~20–35% | Name the risk; active diversification; protect margin on that account |
| Red | Any client above ~35%, or two clients above ~25% each | Treat as a board-level topic: cash, hiring, and pipeline |
Adjust for your cash buffer. A studio with three months of runway can tolerate amber longer than a studio living week to week. A studio with one month of runway should treat amber like red for hiring and fixed costs.
Profit share thresholds can be stricter than revenue share. If one client is 22% of revenue but 38% of gross profit, you are more dependent than the revenue line suggests.
Watch the behavioural signals
Numbers lag behaviour. Concentration becomes dangerous when:
- The big client routinely skips the queue
- Scope expands without change orders
- The team self-censors hard truths in client meetings
- Hiring decisions depend on one renewal
- Forecasts assume the key account renews at the same level without a written basis
- Only one person holds the relationship context
Those are early warnings. They usually appear before the revenue share looks catastrophic. Behavioural signals are especially useful for founders who “know” the account is fine because the last call felt friendly.
Reduce risk without sabotaging the account
Do not punish a healthy client for being large. Diversify around them:
- Protect delivery quality on the key account — resigning through neglect is expensive and damages reputation
- Ring-fence capacity for new logos every month — even 10–15% of senior time reserved for growth prevents total absorption
- Tighten contracts and payment terms while the relationship is calm — not as a threat, as hygiene
- Improve retention discipline on smaller accounts so they stay long enough to rebalance the mix
- Document the relationship — second senior contact, shared brief history, clear escalation path
Acquisition work belongs in Growth. Concentration control belongs in Risk. The same week you celebrate a big win is a good week to name what happens if it ends.
Concentration checklist for leadership
Use this before major hires or lease decisions:
| Question | Yes / No | Notes |
|---|---|---|
| Does any client exceed our amber revenue threshold? | ||
| Does any client exceed our profit threshold even if revenue is lower? | ||
| Would losing our largest client force layoffs within 60 days? | ||
| Do we have at least three months of operating cash without that client? | ||
| Is pipeline coverage realistic for the next two quarters? | ||
| Do two people know each large account well enough to run a transition? | ||
| Are payment terms for the largest client documented and enforced? |
One “yes” on fragility questions is not failure. It is information. The goal is to make the trade-offs visible before they choose for you.
Pair with operational risk
Password ownership, undocumented delivery, and founder dependency compound concentration. A large client plus one person who holds all admin access is a doubled bet. Read them together with the operational risk audit. The full silo is at Risk.
Frequently asked questions
Is a 40% client always a mistake?
Not always — especially early. It is a risk to name, price, and actively diversify against, not a secret to ignore. Many agencies grow through one anchor client. The mistake is building fixed costs as if that share will last forever without measurement.
Should we cap how much we sell to one client?
Often yes. A soft capacity cap forces healthier pricing and protects the rest of the book. Caps can be hours, revenue per quarter, or a rule like “no single client above X% without a written risk plan approved in the monthly review.”
Does insurance fix concentration?
No. Insurance can cover some shocks. It cannot replace lost revenue from a departed key account or recreate the team you built around one retainer.
Is concentration only about losing the client?
No. Concentration also bites when the client stays but shrinks scope, extends payment, or demands more for the same fee. Track profit and behaviour, not only headline revenue.
When should we tell the team?
When concentration affects hiring, bonuses, or role security, transparency helps. You do not need to share every percentage. You do need to avoid surprise layoffs caused by an unspoken dependency everyone sensed but nobody measured.
More agency reading
Share
