Revenue concentration
A year that grew while becoming dependent on eleven households is a worse year than it looks.
Also called Customer concentration, Top-N share, Dependency risk.
What it is
Revenue concentration is the share of income coming from your largest customers, clients or donors. It is a risk measure rather than a performance measure, which is why it is so often absent from a dashboard built to show performance.
revenue from the top N entities ÷ total revenue
- top N
- Usually the top 1, 5 and 10, reported together rather than one alone
- entities
- Resolved customers or households, not records or ordering accounts
- revenue
- One consistent definition, ideally contribution rather than gross
The part the formula leaves out
Identity resolution changes the answer more here than anywhere else, because concentration is a question specifically about the top of the distribution and that is exactly where one entity is most likely to be recorded twice. Two subsidiaries of one group, or two people in one household, look like diversification until the day they leave together.
The second thing missing from the usual version is time. A 52% share held by eleven entities is one risk if the relationships are ten years old and quite another if 14% of it rests on the final instalment of a pledge that ends this year. Concentration without an expiry profile beside it describes the exposure and not the timing.
It should also be measured on contribution rather than revenue where the data allows. A large customer on heavily discounted terms with a high return rate can be a smaller share of profit than of revenue, or a larger one, and only one of those is the risk you care about.
How it is usually computed wrongly
01
Measuring on unresolved records
Splitting one large customer into two mid-sized ones understates concentration precisely where it matters most.
Resolve entities and, where relevant, roll subsidiaries or household members up to the unit that would actually leave together.
02
Reporting one threshold
Top-10 share alone hides whether the risk is one enormous relationship or ten similar ones. Those need completely different responses.
Report the top 1, top 5 and top 10 together. The shape between them is the finding.
03
Ignoring when the money ends
A pledge in its final year and a renewing contract look identical in a share-of-revenue chart, and one of them is about to become a gap.
Put an expiry or renewal date against each large relationship and report concentration by year rather than as a single figure.
04
Treating growing revenue as reducing risk
Concentration and growth frequently rise together, because the fastest way to grow is for a large customer to buy more. The headline improves while the exposure worsens.
Track concentration on the same chart as revenue. Two lines moving together is the signal.
What your file needs
- Transactions with an entity reference and an amount
- Enough identity fields to resolve entities before ranking them
- Contract or pledge end dates, where the exposure is time-limited
Anything missing is reported as unavailable rather than substituted with something weaker computed on worse evidence.
Compute it on your own file
No account needed to start. You only pay when you like what you see.