The Aging Buckets and What Each One Actually Costs

Why this matters

Most shops look at an aging report, see a total, feel bad, and go back to work. It is not a scoreboard, it is a work list where each row already says who should call and what the money is worth by then. Collectability falls steeply with age, so the bucket a balance sits in predicts whether you will see it better than the size of the balance does. Treat every bucket the same way and you spend your collection effort where it is easiest rather than where it is worth the most.

Every bucket carries two readings

The money in a bucket and the number of invoices in it answer different questions, and the money column alone hides half of what the report says.

The money reading is the exposure: what is at risk in that bucket if none of it lands.

The count reading is the cause. One large invoice in the oldest bucket is a customer, and the fix is a phone call to a named person. Nineteen small ones is a process, and no number of phone calls fixes it, because each was individually too small for anyone to bother with. That is how they got old.

A third column sharpens it further: average invoice size in the bucket, as a multiple of your overall average open invoice. Money column over count column, normalized, and it separates the two shapes at a glance.

The decay is steep, and that is the whole reason for a ladder

Recovery-by-age curves published by the commercial collection bar, most often traced to the Commercial Law League of America's long-running collection studies, have shown a consistent SHAPE for decades: near-total recovery while a balance is fresh, still high through the first month past due, falling away sharply through the second and third, a small fraction past a year. That shape is all you should borrow, because the levels vary enough by trade, customer type and residential-versus-commercial mix that anyone else's numbers will mislead you. If you cannot put your hands on the current study, say so and use the measured rates the next paragraph builds in an afternoon.

Computing your own takes an afternoon. Take every invoice issued in a 12-month period ending at least 18 months ago, so the outcomes are settled, find the oldest bucket each one reached, then per bucket add up what was eventually collected against what was billed. Five recovery rates that belong to your shop and your customer mix, re-runnable annually.

The ladder

Boundaries here are measured from the due date, not from the issue date. That matters because the sibling cards on this list use both: a receivable is late relative to when it was due, while the average age of the book is usually measured from issue. Say which one you mean whenever you quote a number.

On work performed on real property, calendar the lien clock separately before any rung below matters, because it does not run on this ladder: mechanic's-lien and preliminary-notice deadlines run from first or last furnishing of labour or materials rather than from the invoice due date, and in many states the preliminary notice falls due within weeks of starting work while the lien filing window closes within a few months of completion. A shop that first calls an attorney at the day-90 rung routinely finds the remedy already expired. Confirm your state's deadlines with counsel at job startup and treat the preliminary notice as a startup task, not a collections step.

One stop rule sits across every rung: the moment a customer disputes the invoice, in writing or verbally and documented, the balance leaves this ladder for whoever owns the relationship until the dispute resolves. A disputed balance does not age, does not go to a credit hold, and above all does not go to a collection agency, because placing a disputed consumer account with a third-party collector is the fastest way to turn a billing argument into a Fair Debt Collection Practices Act claim against the agency and a state unfair-practices claim against you.

One more line has to be drawn before you work the bottom two rungs, because it decides which body of law you are in: a balance owed by a homeowner for work on their own home is a consumer debt, one owed by a business, a general contractor or a property manager is not. The federal Fair Debt Collection Practices Act, 15 USC 1692, binds third-party collectors rather than a shop collecting its own debt in its own name, but it reaches you the moment you write under a name suggesting somebody else is collecting, and several states apply their own conduct rules to original creditors on consumer debts, California's Rosenthal Act at Civil Code 1788 most widely. So on a residential balance, no "final notice from our collections department" letterhead unless a real third party sent it, and check your state before you script the day-30 and day-90 calls.

Those intervals are a starting point for net 30 terms, and the cadence itself belongs to the SOP in the references, AR Aging Analysis and Collections Cadence: that card owns who calls and when, this one owns the recovery weighting, so a reader meeting both runs the SOP's ladder, not two. The six rungs above subdivide the first past-due bucket for ownership while the recovery read below stays on five, and the day-7 rung is a past-due invoice clock, not the seven-day escalation on completed work never invoiced.

Weighting the buckets by what you will actually collect

Shares are of this shop's total open receivable, counts of its 114 open invoices, recovery rates its own from its trailing two years.

The size multiples are derived, not asserted: the average open invoice is 1 over 114, or 0.88 percent of the open money, so the 61-to-90 bucket at 8 percent over 3 invoices is 2.67 percent each, 3.0 times the average, and the over-90 bucket at 7 percent over 19 invoices is 0.37 percent each, 0.4 times the average.

Now weight each bucket by what it recovers:

0.54 times 0.99 is 0.535; 0.22 times 0.97 is 0.213; 0.09 times 0.90 is 0.081; 0.08 times 0.78 is 0.062; 0.07 times 0.35 is 0.025. Those sum to 0.916, so expected recovery is 91.6 percent of the open receivable and the expected shortfall is 8.4 percent of it.

Where that shortfall sits is the finding:

The bucket holding the least money carries more than half the expected loss. Over 90 days is 7 percent of the open money and 54 percent of the expected shortfall, both computed on this same book. Read the money column alone and that bucket is last on the list; read it weighted and it is first by a wide margin.

The second finding decides who makes the calls this week: the 61-to-90 bucket is 1.76 percentage points of the receivable, 21 percent of the shortfall, spread over 3 invoices. That is the highest expected return per phone call on the report, because each call covers an average balance 3.0 times the book's average and each is still recoverable more often than not.

The same shop showing both failure shapes at once

This book has both patterns at once, which is why the count column earns its place. The 61-to-90 bucket worked above is three named customers and three calls: a manager clears it in an afternoon.

The over-90 bucket is 19 invoices, 17 percent of the count, carrying 7 percent of the open money, at 0.4 times average size. Nineteen calls, each on a balance well under average, on money that recovers about a third of the time. Nobody is going to make those calls, which is exactly why those invoices are over 90 days old. Chasing them is not the fix.

That bucket's fix is upstream. Small, old, uncollected invoices almost always come from one of three things: work billed after the relationship went quiet, small balances left after a part payment nobody netted, or invoices sent to a bad address or a former contact. Sample five of the nineteen, find which produced them, fix that, then write the bucket down, because a receivable you will not chase is not a receivable.

Roll the report up by customer before you work it

Aging is produced at invoice level and collection happens at customer level, and the mismatch costs real calls. A customer with rows in three buckets is one conversation, not three: worked row by row you phone them Monday about a day-20 invoice, Wednesday about a day-50 one, and never mention the day-100 row because it sorted to another page.

Here the 114 open invoices belong to 68 customers, and the 19 over 90 days to 11, so the nineteen-call job above is really eleven conversations. Six of those 11 also carry not-yet-due balances, so they are active customers still buying work, and the conversation is about the account, held before the next job is scheduled.

The rollup also reveals concentration the bucket view hides: one customer holding a third of your past-due money across several buckets is a credit exposure problem on one account, not a process problem, and the action is a credit limit and a conversation about terms.

Where the boundaries go for your terms

The 30, 60 and 90 day boundaries are conventional for net 30, where they read as one, two and three times the terms you granted, and they are wrong when your terms differ. On due-on-receipt or net 10 residential work, compress to 15, 30 and 60 days past due: a customer 30 days past due on net 10 is four times past terms, where a net 30 customer sits at day 90. On net 45 or net 60 commercial work the boundaries hold but the day-14 call moves to day 21 and the credit hold with it, because chasing a large account inside its own approval cycle burns goodwill and changes nothing. Retention on progress-billed work is not late and does not belong in these buckets: carry it on its own line with its own release date, or every reading here is wrong in the same direction for as long as the job runs.

References

• Commercial Law League of America collection studies and comparable recovery-by-age data, for the SHAPE of the decay curve only; the levels here are the example shop's own measured rates.
• See related: AR Aging Analysis and Collections Cadence - the SOP that owns who calls and when
• See related: universal-days-sales-outstanding-and-the-mismatch-inside-it, the whole-book figure above this report.
• See related: universal-time-to-invoice-only-counts-the-invoices-you-sent, work that was never billed.
• See related: universal-partly-paid-invoices-and-the-coverage-number, part-paid balances in these buckets.