Customer Credit Limits: Setting the Ceiling and What Happens When It Is Crossed

A business that supplies on credit terms is making a loan without calling it one. A credit limit is the decision about how much it is willing to lend a given customer, and what happens when the debt passes that amount - and at most small businesses both decisions get made in retrospect, in an uncomfortable conversation.

What is actually being decided

A credit limit is not a restriction on the customer but on your exposure. The question is not "how much do they deserve" but "how much could I lose on this customer without it changing my month". That is a question about your business, which is why it can be answered in advance and without new information.

The practical implication: two customers with the same profile can carry different limits at two businesses, and that is correct. What is not correct is a business having no answer at all, so that every large order turns into an argument the owner has with himself.

The three inputs that set the amount

The simple rule that works for most businesses: a limit allowing the number of orders that are open simultaneously under the payment terms, and no more. A customer ordering monthly on net-30 needs a limit of two orders; the same customer on net-60 needs three - and that is already a different risk on identical activity.

What to check before granting credit

Checking beforehand is the cheapest action in the whole life cycle of the debt. What is actually examined:

• The identity of the legal entity placing the order, not the brand name.
• How long that entity has been active.
• Whether there is payment history with you, and what it says.
• Who is authorised to order on the customer's behalf.
• An address for documents and a contact for accounts.
• Whether the first order is on prepayment.

The first item is where businesses fall. An order arrives under a trading name, the invoice is issued to a different entity, and when it is time to collect it turns out to be unclear who exactly the agreement is with. Pulling the entity's details from the companies registry is a short check, covered in depth in checking a customer against the companies registry.

What happens at the moment of excess

This is where policy fails. At most businesses an excess raises a message in the system, and then the order ships anyway - because the customer is waiting and the warehouse does not decide credit.

A policy that works defines three tiers in advance, each with an owner:

1. Up to the limit - approved automatically, nothing to decide.
2. A small excess - approved by whoever owns the relationship, and recorded.
3. A material excess, or debt in arrears - halts supply until the owner decides.

What makes this work is that the second tier exists. Without it only two options remain - approve everything or stop everything - so in practice everything gets approved.

Why a long-standing customer who is late is a separate case

The natural instinct is leniency toward a long-standing customer, and it is reasonable: there is history, there is trust, and the relationship is worth more than a single order. The problem is that lateness at a long-standing customer is usually the only early sign that something has changed at their end.

So the operational rule is not to stop but to ask. A long-standing customer running two months late after two years of precision is not a collections case - he is a conversation. What you cannot do is keep supplying quietly and assume it will sort itself out, because when it does not the exposure is already three times larger. The link between that signal and the collection process is in an orderly receivables collection workflow.

Three routes when a new customer asks for credit

No binary decision between yes and no is needed. Three routes cover nearly every case, and each lowers exposure without losing the deal.

The first is prepayment on the first order only, moving to credit terms after two or three deals paid on time. It sounds rigid and is accepted easily in practice, because it is presented as a procedure rather than as distrust.

The second is part payment up front - some with the order, the balance on terms - and it is the common route on large orders. It cuts exposure by roughly the amount covering direct cost, which is usually what matters.

The third is full credit from day one, and it makes sense when the up-front check is strong: an established entity, clear activity, a named accounts contact. What matters is that the route be chosen in advance and recorded on the customer record, so the next order does not reopen the same discussion.

Where the limit has to live

Somewhere it is visible before an order is approved - that is, on the customer record, not in a policy document. A limit appearing only in a file nobody opens is not a policy but an intention.

What is actually needed is three fields on the customer record: the limit, the current open debt, and the date the limit was set or last updated. The third sounds redundant and it prevents the common situation in which a limit set when the customer ordered small amounts stays in place three years after the volume changed entirely.

How often should a limit be revisited?

A year is a reasonable cadence for most businesses, and beyond that two triggers justify an immediate update: a material change in order volume, and a first late payment. Both change the assumption the original limit rested on, and both are easy to spot if the limit sits next to the open debt.

What you should not do is update a limit in the middle of negotiating an order. At that moment commercial pressure is at its highest and information at its lowest, and the decision gets made by the need to close a deal rather than by the exposure.

Sources

• Israel Corporations Authority - company information
• Checking a customer against the companies registry
• An orderly receivables collection workflow
• Book a call