Reading a receivables ageing report without guessing
An ageing report is a claim about time, not a column of dates. What each band actually asserts, why a total outstanding figure hides cannot-pay from will-not-pay, the two questions to ask per bucket, and what the report can never tell you.
An ageing report is a claim about time, and it is the most misread page in small-business finance. The mistake is usually one of two: treating the oldest bucket as the urgent one, which is often reasonable and occasionally absurd — a single ancient invoice at a client you will never bill again is not the same problem as a quarter of current invoices at your best client — or treating the total outstanding as a collection problem, when at least as often it is a billing or pricing problem that no amount of chasing will touch.
Both mistakes come from reading the buckets as a sort order rather than as assertions about promises. This is the companion to our piece on why an invoice and a payment are different records, and the connection is not decorative: an ageing report is not stored anywhere. It is computed from payment allocations each time you open it, so its trustworthiness is exactly the trustworthiness of the records underneath it.
An invoice is not overdue because of the date on it
The date on an invoice says when the document was issued. Overdue is a relationship between that date and a promise: the terms the client was given and the date payment was expected. An invoice with 30-day terms issued on the first is not late on the twentieth, however old the row looks in a list. A 60-day client at 45 days is not a collections problem, and any report that flags them is measuring your terms table rather than their behaviour. So the question per bucket is not how old is this but what was promised, and when. Our free receivables ageing calculator puts rows into 0–15, 16–30, 31–60, 61–90, and 90+ bands measured from the invoice date, which is the right simplification for a spreadsheet typed in a hurry — and the page says plainly that it does not handle credit terms. In the operating record the terms travel with the quote and the invoice, so a band can be read against the promise that client actually received.
What each band actually asserts (a structural reading, not a statement about your customers)
| Band | What it asserts about the promise | The honest first question |
|---|---|---|
| 0–15 days | Nothing has failed yet, or the promise is not yet due | Is this client inside the terms they were given? |
| 16–30 days | The promise is past due, or just due, depending on the terms record | What date did we give this client, in writing? |
| 31–60 days | The promise was missed once and no reason was recorded | Is there a recorded promise, or only silence? |
| 61–90 days | What was said and what happened have stopped matching | Who owns this conversation now? |
| 90+ days | These promises are usually abandoned, not merely delayed | Why has this one stopped being anybody’s job? |
A total outstanding figure hides the difference between cannot pay and will not pay
A total is an average of unrelated situations, and averaging is the operation that destroys the decision. The same figure can be made up of a client waiting on a purchase order that arrives next week, a client in a payment dispute over one line, and a client you stopped calling months ago. Three different actions, and the total tells you only that money exists somewhere. It also hides something worse: a total is a collection figure by assumption, framing the problem as people who have not paid, when a large part of outstanding is often work that was delivered and never invoiced — and ageing cannot see that money, because there is no invoice for it to age. A total also hides concentration: the same amount spread thinly across many small clients is a process problem, while concentrated in one relationship it is a commercial decision.
Read the buckets beside the four money states, not on their own
Ageing is a view on one of four states — quoted, billed, collected, and outstanding — specifically on the gap between billed and collected. On its own it tells you about collection. Beside the others it tells you more. Billed falling short of quoted means work was agreed and delivered without a document, which is a billing gap, and chasing the client will never surface it. Outstanding growing while billed stays flat says something about terms, capacity, or how hard work is being pushed out of the door. And a client can look healthy on every state except collection, which is worth knowing before the next proposal goes out. Resolution matters too: per client tells you who to call, per project tells you whether you billed for the work you actually did.
The two questions to ask per bucket
Two questions, and nothing else is worth arguing about first
Who promised what
Which line items the client accepted, on which quote, with which tax treatment. A row with no agreed scope behind it is an amount, not a claim: you can chase it, but not defend it when the amount is questioned. This is also where an unbilled addition shows up — a gap, not a receivable.
When did they promise to pay
The promised date, not the invoice date. A recorded promise turns the follow-up from a chase into a specific question: did something happen between the promise and now? Where none was recorded, the oldest bucket is measuring your follow-up discipline, not the client’s behaviour.
What to have in front of you before you pick up the phone
Three things, and the fourth is usually the answer
- The accepted quote behind the invoice, so you know what was agreed rather than what was later assumed
- The invoice date and the terms on that quote, so “overdue” refers to something
- The recorded payment promise and its date, if one exists — the specific thing to refer to
- Whether the invoice is fully outstanding or partly allocated, and by how much
Reading a bucket honestly
Two columns, no middle ground. A report you can act on tends to be strong on all of these.
Why the report can only be as good as its allocations
This is where the companion article earns its place, so we link the thinking rather than repeat it: the invoice’s paid state is not a stored flag, it is a projection of payment allocations. Outstanding is the invoice total minus the sum of allocations, recomputed rather than maintained, and ageing is computed from that same projection every time you open the report — the honest reason to treat it as a view rather than a record. Four things distort it. An unallocated payment first: money arrives with no remittance advice and covers four invoices, and until somebody decides, the invoices read as fully outstanding while the cash sits in the bank, with the oldest of them looking like the worst collection failure in the business. An allocation dated later than the payment changes which band an invoice sits in at the moment you look, because ageing reflects when money was applied, not when it arrived. A credit note reduces the claimable amount and belongs in the same arithmetic. And late entry moves invoices between bands after the fact, which is why a report without an as-of moment cannot be compared with last month’s.
What an ageing report cannot tell you
It cannot tell you whether a client will pay. It is a statement about elapsed time, not about intent, and no band predicts behaviour. It cannot see work that was delivered and never invoiced, so the most commonly missed loss in the chain is invisible in it by construction. It cannot tell you whether the work behind the invoice was worth doing — that is a project economics question about quoted against billed and the costs you actually record, and a profitable client can sit at the top of your ageing table without irony. It is not a statutory ageing statement: no ledger, no trial balance, no period close. And it does not chase anyone. Collection needs a named owner, and a shared outstanding list with nobody accountable ages every month until the oldest items are too awkward to raise — which is exactly how they end up written off.
What we have not measured
We have no measured collection rate, no measured days-sales-outstanding figure, no measured ageing distribution, and no customer data, and we present none of those. The worked example in our free ageing calculator is arithmetic on placeholder numbers to show how bands behave, not a sample of a real receivables book. Our ageing is an operational read of your own invoices and payment promises, not a statutory figure, and the two should not be quoted interchangeably — if you need the second, your accountant computes it from the ledger. What we would want to count before making any claim about collection behaviour: the share of overdue invoices carrying a recorded promise, the time between a promise and the money arriving, and how outstanding moves against billed over several quarters.
Frequently asked questions
What does each ageing bucket actually mean?
A statement about elapsed time since a promise. The 0–15 band asserts that nothing has failed yet, 31–60 asserts that the promise was missed once with no reason recorded, and 90+ usually means those promises have been abandoned rather than merely delayed. The bands only carry that meaning next to the terms the client was actually given, because overdue is a relationship between a date and a promise, not a property of the date.
Should I always chase the oldest invoice first?
No. It is a reasonable starting point and a bad rule. Oldest first within a client is useful; oldest across your whole book usually means an abandoned client relationship competing with a current one. Size the relationship next to the age before you decide, and remember that a large portion of outstanding is often unbilled work that chasing the client will never surface.
Is a high total outstanding a collection problem?
It might be, and it is at least as often a billing or pricing problem. If billed is falling short of quoted, money is sitting in delivered work that was never invoiced, and no follow-up call will find it. Read the total against billed and quoted, per client, before treating it as a collections issue.
How should a part-paid invoice appear in ageing?
On the outstanding amount rather than the gross, because the invoice’s paid state is a projection of payment allocations, not a stored flag. Anything still unallocated keeps the invoice looking outstanding while the cash is already in the bank, which is the most common way an ageing report overstates a problem.
Does the ageing report replace my accountant’s ageing statement?
No. It is an operational read of your invoices and payment promises, not a statutory ageing computed from the ledger under an accounting standard. The two are not interchangeable, and if you need the statutory version your chartered accountant produces it from the books.
What if there is no payment promise recorded?
Then the oldest bucket is measuring your own follow-up discipline rather than the client’s behaviour, and it is worth knowing that before you attribute the delay to them. The practical fix is to record a promised payment date whenever one is given, so the next conversation can refer to a specific commitment rather than to silence.