Owed £22,000 at any given time: turning late-payment chaos into a chase list

The UK's late-payment problem is structural. Around half of SME invoices are overdue at any given time; the average small business waits roughly a month beyond terms to be paid and is owed about £22,000 at any moment. Across the economy, late payments are estimated to cost close to £11 billion a year and to contribute to tens of thousands of business closures. The government's late-payment reforms will tighten reporting duties on large buyers — but no legislation collects your invoices.

Chasing works; the follow-up call is the most effective collections tool an SME has. But you can only chase what you can see — and in many small businesses, the honest answer to "which invoices are unpaid right now?" is a spreadsheet that was accurate two weeks ago, maintained by someone who has other jobs.

Why the unpaid list is always stale

The unpaid list decays because it is assembled by hand from three places that never talk to each other:

• Invoices live in one system (or a folder of PDFs)
• The bank statement lives in another
• The matching between the two happens in someone's head, usually at month-end

Between two reconciliations, every list is a guess. And a stale list is worse than no list: chase a client who paid last Tuesday and you look disorganized; fail to chase one who has gone quiet for 45 days and you finance their business interest-free.

The psychology compounds the mechanics. Chasing feels confrontational, so it gets postponed; postponement makes the eventual conversation vaguer ("I think there are a couple outstanding?"), which makes it easier for the debtor to stall. Vagueness is the late payer's best friend.

The mechanical fix: reconciliation as a by-product

When bank statements are reconciled against invoices automatically — on import, not at month-end — the unpaid list stops being a chore and becomes a by-product:

• Import the statement: CSV from Lloyds, HSBC, Barclays, NatWest or Starling, or OFX — the format and each bank's column quirks are detected automatically
• Each payment is matched to its invoice on amount, date, reference (including references buried in payment memos) and counterparty — including the awkward case of one transfer settling several invoices
• Confirmed matches are written onto the entries: settled invoices leave the list for good, with the statement line recorded as evidence

What remains, permanently current, are exactly two lists. Payments with no invoice: bank fees, surprises, the occasional mystery worth a look. And invoices with no payment — which is your chase list, with due dates extracted from the invoices themselves, so it sorts by urgency without anyone maintaining it.

From list to action

A current list changes the conversation. Compare:

• "I think you might owe us something from the spring?" — deniable, stallable, forgettable
• "Invoices 0182 and 0197, due on the 14th, remain unpaid — shall I resend the copies?" — precise, evidenced, awkward to dodge

Every invoice in DOXALIO carries its extracted payment terms and due date, and proactive alerts surface documents whose deadline is approaching or has passed — so the chase starts on day one of lateness, not at the month-end reckoning. Page-cited evidence (the invoice, its terms, its due date) attaches to every claim you make.

Consistency beats aggression. A supplier who invoices cleanly, reconciles daily and follows up on day one — politely, every time — migrates to the top of clients' payment queues. Late payers triage their creditors; the precise ones get paid first.

The cash-flow arithmetic

Cutting average collection time by even ten days on £22,000 of standing receivables is permanent working capital returned to the business — roughly £600 per £22k at current borrowing costs, every year, plus the closures-grade risk reduction that no spreadsheet line captures. The tooling cost of maintaining the list automatically rounds to nothing against either number.

FAQ

Does automated reconciliation work with UK bank CSV exports?

Yes — that is most of the engineering. Lloyds and HSBC split money-out and money-in columns, NatWest uses a signed Value column, Starling labels counterparties differently; the importer recognizes each dialect (plus OFX) and normalizes before matching.

What about clients who pay several invoices with one transfer?

That grouped payment is the case that breaks amount-only matching — no single invoice equals the transfer. The matching engine searches invoice combinations against the transaction and proposes the group with a confidence score; one click confirms all of them.

Is this a debt-collection service?

No — it is the information layer that makes your own chasing effective: a permanently current list of who owes what, since when, with evidence. What you do with it (reminders, statutory interest under the Late Payment Act, escalation) stays your call, made on facts.