On Monday morning the collections team opens the aging report, starts in the over-90 column and works left. By Wednesday they reach the 31-to-60 bucket. Somewhere in it is the company's third-largest customer, who has paid on time for two years and slipped to 45 days last week. Nobody has called them yet.
That is collections prioritization by default. It is orderly and easy to audit, and it puts the most effort on balances that are least likely to move. The aging report is a good record of what is owed. As a call list it has serious gaps.
Why the aging report is a poor collections prioritization tool
It ranks by time, not money. Take a $900 invoice at 120 days and a $400,000 balance at 40 days past due. Bucket order puts the first ahead of the second.
Old items are often not collection problems. The oldest buckets fill up with disputes, deductions, pricing errors and unapplied cash. A phone call will not clear any of them. They need an owner and a fix.
It hides change. A customer who moves from current to 45 days for the first time is telling you something. A customer who has sat at 100 days for six months is telling you something different. Both look like a number in a bucket.
It often ages from the wrong date. Many aging reports count days from the invoice date. A customer on 60-day terms at 45 days is not late, and a customer on 15-day terms at 40 days is very late.
It treats every customer alike. A long-standing account with a large forward order book and a one-off buyer with no further business get the same treatment if their balances sit in the same column.
None of this means the aging report is wrong. It answers a different question. The board's DSO figure is built from the same balances, and that is why aging matters for reporting. For reference, the median DSO in APQC Open Standards Benchmarking (cross-industry median, all companies) is 38.0 days, across 13,195 companies. For deciding who to call, it needs other inputs.
What to rank on instead
| Factor | What it tells you | Where the data usually lives |
|---|---|---|
| Amount past due, from the due date | What a successful contact is worth | Receivables ledger and payment terms |
| Change in payment behavior | Whether a reliable payer is slipping | Payment history |
| Promise history | Whether this customer's commitments can be relied on | Collector notes, often in email |
| Dispute status | Whether a call can collect anything at all | Dispute log, deductions, customer service |
| Credit risk | How much is exposed if they stop paying | Credit file, limits, external reports |
| Customer importance | What a heavy-handed call might cost | Sales, contracts, forward orders |
The first three push an account up the list. Dispute status takes it off the call list and onto someone else's. Credit risk and customer importance change how you call, not only when. A large, important customer who is slipping warrants a call from a senior person, possibly with the account manager, rather than a third reminder email.
Keep the scoring simple and visible. A collector who can see why an account ranks where it does will trust the list. A black-box score gets ignored by the second week.
A worklist method
1. Take out what a call will not fix. Every morning, route disputed items, open deductions and unapplied cash to their owners. They leave the call list but stay on a separate list with a name and an age against each.
2. Rank the rest. Score each remaining account on amount past due, any change in payment behavior and broken promises. Sort by score, not by bucket.
3. Set the action by tier. The top of the list gets a call today. The middle gets a personal email. Small, low-risk balances get an automated reminder. Large accounts with credit concerns go to the credit manager for a decision on holds.
4. Record the outcome. A promised date and amount, a new dispute, a request for a copy invoice, or no answer. Each outcome sets the next action and its date.
5. Feed promises back in. A promise kept lowers the account's priority next cycle. A promise broken raises it. Over a few months, promise history becomes one of the most telling inputs you have.
Size the list to what the team can do in a day. A list nobody finishes trains people to ignore the bottom half.
Where AI helps in collections
Scoring every day. Recalculating the ranking from the ledger, payment history and collector notes each morning is tedious by hand and straightforward for software. AI is useful where the inputs are unstructured, such as reading notes and emails to find promises and disputes.
Summarizing an account before a call. A short summary of open items, recent payments, promises and disputes spares the collector a tour of several screens before each call.
Reading replies. Customer emails can be classified as a promise to pay, a dispute, a remittance or a request for documents, and routed accordingly.
Drafting reminders. AI can draft a reminder that reflects the account's actual position. A person reviews it and sends it.
What stays with people: credit holds, payment plans, write-off proposals and any conversation where the relationship matters more than the invoice. Watch for one specific failure. If the score is trained on how collectors worked in the past, it can learn the aging report's habits and send effort back to the oldest buckets. Check the ranking against outcomes, not against what the team used to do.
What a ranked worklist will not fix
It will not fix invoices that are wrong when they leave. If customers cannot match an invoice to a purchase order, or the price differs from the contract, collection is slow whatever the ranking says. The cure is in billing.
It will not fix credit decisions already made. A ranked list tells you who to call first. It does not make a customer who was given too much credit able to pay, and terms agreed in a sales negotiation stay agreed until someone renegotiates them.
Frequently asked questions
What is collections prioritization?
Collections prioritization is the daily decision about which customers to contact first, based on the amount at stake, the risk of non-payment and the likelihood that a contact will produce cash.
Why is the aging report not enough to prioritize collections?
The aging report ranks balances by time alone. It ignores amount, payment trend, promise history and disputes, so it tends to focus effort on old, small or disputed items while new risks in large accounts wait.
What factors should be used to prioritize collections?
Amount past due from the due date, changes in payment behavior, promise history, dispute status, credit risk and customer importance. Disputed items should leave the call list and go to an owner who can resolve them.
How often should a collections worklist be updated?
Daily for the ranking, because payments, promises and disputes change every day. Review the method itself monthly against what was actually collected.
Can AI prioritize collections?
AI can rank accounts daily from ledger data and collector notes, summarize an account before a call and draft reminders for review. Credit holds, payment plans and sensitive customer conversations should stay with people.
Where CFOLogic fits
For collections, our approach is a worklist that reads from the ERP without changing it, ranks the accounts that need a person, and shows who owns each item and how long it has waited, with reminders drafted for a collector to review and send. We would normally test it on one team or business unit before going wider. It sits alongside the finance operations support described in Finance Team as a Service.