Most business failures are visible long before they are announced. In more than 35 years in this industry I have rarely seen a customer go from reliable payer to bad debt in a single step. The warning signs are usually there months earlier, sitting in data the supplier already holds or could easily get.
Atradius published a thoughtful piece last month on the customer risk signals every CFO should be watching. I agree with almost all of it. Payment behaviour, receivables ageing, credit utilisation, ordering patterns and portfolio concentration are where trouble tends to surface first.
Where I would take the argument further is on the source of those signals. Every one of them comes from your own sales ledger. That makes them valuable. It also means they only ever show you half the picture.
What your own ledger can tell you
The sales ledger is one of the most underused sources of risk intelligence in any business. Most finance teams use it to chase cash. Far fewer use it to read how their customers are changing.
That trend is something to look at. One late invoice is usually administrative. A customer drifting from 30 days to 45, then to 60, over the course of a year is telling you something about its cash position. And it’s worth paying attention.
The signals always become much stronger when they line up. A customer running close to its credit limit may simply be growing. The same customer running close to its limit while paying more slowly and placing smaller, more frequent orders is far more likely to be protecting its cash.
Concentration deserves separate attention. A large customer with a clean payment record can still do more damage than a weak one if it accounts for a fifth of your receivables. The size of an exposure and the quality of the customer behind it are two different risks, and I would always measure them separately.
What your ledger cannot see
The limitation is simple. Your ledger only records how a customer treats you.
Businesses under pressure are rarely late with everyone at once. They prioritise. The suppliers they cannot operate without get paid on time, often right up to the end, and everyone else waits. If you happen to be one of the protected suppliers, your ledger will look perfectly healthy while the customer stretches HMRC, its landlord and its smaller creditors.
If you happen to be one of the protected suppliers, your ledger will look perfectly healthy while the customer stretches everyone else.
There is a timing problem as well. Ledger signals need history to mean anything. A new customer, or one you only trade with occasionally, gives you very little to read, and that is often exactly where fresh exposure is being taken on.
The public record fills that gap. Late filing of accounts, a weakening balance sheet, county court judgments, winding up petitions, new charges registered against assets and a run of director resignations will never appear in your ledger. Some of them can surface well before a customer starts paying you late.
The headline numbers hide the pressure
On the surface, the latest official figures look reassuring. The Insolvency Service recorded 1,882 company insolvencies in England and Wales in September 2026, 5% fewer than a year earlier. Underneath that, compulsory liquidations were 25% higher than in September 2025.
That mix matters. A creditors’ voluntary liquidation is usually a decision taken by the directors. A compulsory liquidation follows a winding up petition, which typically means a creditor has run out of patience and gone to court. A rise in those points to creditors enforcing harder, and every one of those petitions sat on the public record before the order was made.
The overall rate is still significant. One in 200 companies in England and Wales entered insolvency in the 12 months to the end of September. Across a portfolio of a few thousand trade customers, that is a meaningful number of failures to see coming.
Reading both sides together
The earliest and most reliable warning comes when the inside view and the outside view are read side by side. A customer paying you a little more slowly is a prompt to take a closer look. A customer paying you more slowly while its probability of distress is climbing needs a decision today.
A customer paying you more slowly while its probability of distress is climbing needs a decision today.
That is the thinking behind PoD®, the Probability of Distress score we launched in October and now the primary Company Watch score for distress. PoD® puts a single probability on whether a business is heading into difficulty, so a credit team can rank every customer in a portfolio and see straight away where attention is needed first. A rising PoD® on a customer whose ledger still looks clean is exactly the blind spot I described earlier, brought into view.
The H-Score® still sits alongside it as a measure of financial health from filed accounts. We show the workings behind every score, because a number you cannot explain is a number you will struggle to act on.
Credit and risk teams have more data available to them than ever. The ones that avoid losses will be those that connect what their own ledger is telling them with what the rest of the market can already see, early enough to reduce exposure.
If you’d like to see how PoD® reads across your own customer book, book a demo today.
Payment behaviour, receivables ageing, credit utilisation, ordering patterns and portfolio concentration are where trouble tends to surface first. The trend matters more than any single invoice: a customer drifting from 30 to 60 days over a year is telling you something about its cash position, especially when it is also running close to its credit limit or placing smaller, more frequent orders.
Your ledger only records how a customer treats you. Businesses under pressure prioritise the suppliers they cannot operate without, so a protected supplier can see a clean payment record while the customer stretches HMRC, its landlord and smaller creditors. New or occasional customers also have too little history to read. Public record signals such as late filing, county court judgments, winding up petitions, new charges and director resignations fill that gap.
The Insolvency Service recorded 1,882 company insolvencies in England and Wales in September 2026, 5% fewer than a year earlier. Compulsory liquidations, which follow a creditor’s winding up petition, were 25% higher than in September 2025, and one in 200 companies entered insolvency in the 12 months to the end of September.
Read the inside view and the outside view side by side. Company Watch’s PoD® puts a single probability on whether a business is heading into difficulty, so you can rank every customer and see where attention is needed first, while the H-Score® measures financial health from filed accounts. A rising PoD® on a customer whose ledger still looks clean is the blind spot brought into view.
Craig Evans
CEO
It’s the people and clients that make our business, and at Company Watch we are fortunate to have both exceptionally talented people and an extensive set of loyal clients. I feel very privileged to lead the company and partner with our customers in this next exciting chapter.
