A Pound Well Spent: Screening New Trade Customers Before You Extend Real Credit

Most credit decisions taken by small UK suppliers rest on a familiar ritual: a Companies House check, a credit score, two trade references that were chosen by the customer, and a phone call to an account manager who is paid to be positive. None of it tells you how the business actually behaves when an invoice lands in its purchase ledger.

There is a cheaper way to find out. Before you extend real credit, sell something small and observe what happens. The value of the transaction is almost irrelevant; the behaviour it exposes is not.

Why a nominal invoice outperforms a credit check

A credit report is a historical document assembled from filed accounts, county court judgments and lender data. It says nothing about whether your customer's accounts payable function requires a purchase order number that nobody told you about, whether invoices are approved monthly by a committee, or whether the finance team treats a 30-day term as a suggestion.

A small paid transaction tests the live process. You learn the length of the approval chain, the quality of communication from the ledger team, and — most valuably — the gap between the due date on your invoice and the date money actually arrives. That gap is the single most useful number in credit management, and you can obtain it for the price of a sandwich.

Designing the test so it counts as genuine trade

There is an important discipline here: the transaction must be a real supply on real terms, not a contrived invoice. Issue a genuine VAT invoice for a genuine deliverable at a genuine price. Suggested forms include:

• a paid sample or trial batch of goods;
• a short paid consultancy or scoping session;
• a nominal set-up or onboarding fee for a service;
• a first order deliberately sized below your usual minimum.

Set the price low enough that losing it would not matter, but high enough to be plausible. One pound is symbolic; five to fifty pounds is usually more informative, because a sum that small can escape the standard payment run altogether. That is the first limitation to understand: low-value invoices are sometimes settled through petty cash or a weekly sundries payment, which can make a slow payer look quick, or a quick payer look slow. If you want a cleaner reading, run the test at a value that must pass through the normal ledger — and, where practical, repeat it once.

Make the terms explicit in writing at the point of order: 14 or 30 days, payment by bank transfer, with a named contact for queries. Note the due date on the invoice itself, not only in the covering email.

What to record while the test runs

Keep a simple, dated log from the moment the order is placed. The following observations are worth capturing.

Ordering friction. How long did it take to obtain a purchase order? Was a supplier number required, and how many forms, certificates and bank letters did onboarding demand? A forty-page vendor questionnaire for a twenty-pound transaction is not a red flag in itself, but it is a reliable indicator of a heavily bureaucratised ledger.

Approval speed. How many days between delivery and invoice approval? Did anyone confirm receipt? Did the invoice have to be resubmitted in a different format?

Query behaviour. Were trivial items disputed? Did the customer challenge VAT, carriage or a line description that had been agreed in writing? Suppliers who query pennies on a small invoice frequently query pounds on a large one, and each query restarts the clock.

Communication quality. Did you get a remittance advice? Was the named contact reachable and did they answer questions directly, or did correspondence disappear into a shared inbox?

The payment gap. Record the invoice date, the due date and the actual date funds cleared. Then note the number of days between each.

Payment mechanics. Did the money arrive from the customer's own account, from a group treasury entity, or from a third party? Was it paid in a single transfer or split across runs?

Reading the result

A simple scoring frame helps turn observations into a decision. Score each of four dimensions from one to five: approval speed, paperwork friction (inverted, so five means minimal), payment against due date, and communication quality. A total near the top of the range supports standard terms. A middling score suggests conditions — a deposit, shorter terms, staged payments on larger orders, or a credit limit set at one month's expected spend.

A low score is information, not an insult. It tells you the price of doing business with this customer includes an extra thirty or sixty days of working capital, and that cost belongs in your quotation, not in your overdraft.

When the test should change the terms

Treat the following as prompts to tighten rather than proceed. A customer who pays a nominal invoice more than thirty days beyond its due date has demonstrated, unambiguously, how it will treat a five-figure invoice. One that cannot identify a finance contact, or insists that terms be renegotiated to 60 or 90 days before the first transaction has even completed, is telling you where the risk sits. Persistent refusal to issue a purchase order before delivery, demands for retrospective credit, or a pattern of disputes over immaterial sums all point the same way.

In those cases, consider pro-forma payment for the first three orders, a deposit with the balance on delivery, or payment by direct debit where the relationship justifies it. If the customer resists every one of those options while still pressing for open credit, you have your answer — and you have it before the exposure exists, which is the entire point.

The limits of a pound

Be honest about what a single small transaction cannot prove. It samples one route through the ledger, at one moment, handled by one clerk. Behaviour changes when sums get larger and when the customer's own cash position tightens. Approval chains that were brisk in a quiet quarter can lengthen markedly at year end.

That is why the test is best used as one input among several, alongside filed accounts, payment practices reporting where available, trade references you have chosen yourself, and a credit limit that reflects your own capacity to absorb a bad debt. It is not a guarantee. It is a very cheap way of replacing a guess with an observation — and in credit management, that is usually the best return on a pound you will find.