Credit Limit Setting for New B2B Customers

Setting credit limits for new B2B customers means deciding how much exposure to accept before the buyer has paid a single invoice. Finance teams need enough evidence to protect cash without turning away buyers who can pay, plus clear rules for changing the limit as they learn how each buyer pays.

Why the first credit limit for a new B2B customer is structurally harder than any later decision

Later credit reviews have internal payment history to draw on. The opening decision has no such advantage. It rests entirely on external evidence, which can indicate whether a customer is likely to pay but cannot show how that customer pays this seller. A credit limit caps financial exposure before payment is confirmed, so treating it as a favor to a new buyer turns a decision about risk into a welcome gift.

The stakes also pull in opposite directions. After a write-off, new revenue must bring in enough margin to cover what was lost, which takes more than a sale equal to the unpaid invoice. Yet a limit that is too cautious can shrink orders and send buyers who can pay to a competitor willing to offer terms. Allianz Trade’s credit limit guidance describes this tension as the need to protect cash flow while leaving room for growth. A structured process resolves that tension more reliably than either gut feel or a single reassuring score.

What a written credit policy must settle before any individual limit is set

A written policy establishes risk appetite and approval authority before an application arrives. Without it, each limit becomes a one-off negotiation, making decisions inconsistent and difficult to enforce fairly. The policy must define the maximum loss the business will accept from one account as a share of revenue. It must also name who can approve each dollar tier and what documentation that authority requires.

Starting terms belong in the same policy, including payment terms, discounts and deposit requirements. So do the events that trigger a review or hold, and the escalation path when a decision exceeds someone’s authority. A credit analyst should not have to discover who can approve an exception while a salesperson waits for an answer. The authority to say yes needs to be as clear as the authority to say no.

Sales and finance must both sign off before the policy goes into use. Otherwise, credit control may get the blame for standing between a buyer and an order. The shared case is straightforward. Unchecked accounts can produce write-offs that erode the revenue sales worked to win, while clear approval rules create a consistent route for viable customers.

Gathering the inputs before setting a number

The policy’s dollar thresholds should determine how much evidence each application needs. For a small limit, a bureau report and two trade references provide the prescribed starting check. A request in a large-limit tier warrants financial statements and deeper diligence. The dividing line belongs in the policy. Applying large-account scrutiny to every small buyer creates a bottleneck without respecting the size of the exposure.

Within those tiers, the minimum evidence spans financial data, third-party payment behavior and business background. Commercial reports from Dun & Bradstreet, Experian Business and Equifax provide payment information and public records, including liens. Bureau payment index scores offer a cross-supplier view that selected trade references cannot provide. References are useful, but applicants choose them. A glowing reference is less persuasive when the applicant also chose the spotlight.

For larger limits, collect two years of profit and loss statements and balance sheets. Debt-to-equity and cash flow compared with what the buyer owes help show whether it can pay, while bank references show where it holds deposits and how old its accounts are. Business background adds years in operation, ownership structure and industry conditions. Sector stress can change the reading of otherwise reassuring company information. No single source answers both whether the buyer can pay and whether it tends to pay.

The application process gives further signals, including how fast buyers respond, how hard they push back on standard terms and whether first orders are unusually large for their stated company size. These signals need to be weighed before a buyer is turned down. Expected sales volume also belongs in the file because the proposed limit must fit the intended relationship. The guidance lists that volume, payment history, creditworthiness, reports, financial statements and industry risk as the main inputs.

Translating inputs into a defensible dollar figure

A repeatable scoring method turns the file into a risk assessment. Add points for positive factors such as five or more years in business, a clean credit report, a strong bureau payment index and established trade references. Deduct points for a new business, slow payments, legal judgments or high debt-to-equity. Keep the model simple enough to use the same way each time. A complicated score can still be a guess wearing a tie.

Under the policy, applicants with high scores get full terms, those with mid-range scores get a lower limit, and those with low scores must use cash on delivery or prepayment. Those thresholds must reflect the business’s stated risk appetite rather than an analyst’s comfort level on a particular day. The score sorts the risk, but it does not yet establish an affordable dollar exposure.

Two calculations provide the next boundaries. The seller-side cap equals the seller’s monthly revenue multiplied by the percentage allowed under its policy. The buyer-side cap is the customer’s annual revenue multiplied by a small percentage that is allowed. Each calculation produces a dollar amount from a different question. The first asks what the seller can afford to expose, while the second asks whether the extension is proportionate to the buyer’s turnover.

The proposed limit must stay within both boundaries and take the score into account. A buyer’s substantial revenue does not give the seller a larger capacity to absorb losses, and a financially strong seller does not make a weak buyer safer. The percentages therefore need to come from policy. For customers with limited history, the guidance recommends conservative starting limits, with increases only after consistent on-time payments.

Before approval, calculate the buyer’s share of total accounts receivable if it draws the full proposed limit. Concentration can defeat an otherwise sound individual assessment. If the share would be significant relative to the portfolio, tighten the initial limit and increase review frequency. Payment terms should graduate with the limit, with new or unverified buyers starting on cash in advance or shorter terms for the first few orders before moving to standard net terms after demonstrated reliability. The amount and the time allowed to pay are parts of the same exposure decision.

Where human judgment still sits in a process that can be partly automated

Automation can gather bureau data, references and financial signals, apply scoring criteria the same way each time and flag applications for fast-track approval or manual review. For standard accounts, it can reduce approval time from days to hours. The binding decision to approve an account, assign terms and create available credit still requires an authorized person.

People still need to use their judgment when parts of the file are missing or the legal-entity match is unclear. Someone must decide whether more documentation is needed and whether an exception is justified. Approved exceptions must be logged formally so they can be reviewed if the account goes bad. This division avoids both extremes. Teams need not preserve manual data gathering to prove they are careful, or remove human sign-off to prove they are modern.

Credit assessment should run in parallel with contract negotiation. Approval must remain separate from the sales relationship, and any decision made over the credit team’s objection must be formally documented. European operations also need to check which regulations apply. The European Commission states that the EU AI Act entered into force on 1 August 2024 and identifies AI used to assess the creditworthiness of natural persons as high-risk, rather than classifying every corporate credit assessment that way. Where that classification applies, data governance must address representative and nondiscriminatory training data, so compliance work belongs in the plan.

How trade credit insurance changes both the limit and the information available

A trade credit insurer assigns a buyer limit defining the maximum outstanding receivable it will cover. That figure also provides credit intelligence. It can anchor or cross-check the seller’s calculation because the insurer brings information the seller could not generate alone. If coverage is only a fraction of the proposed exposure, the gap warrants review.

A reduction or withdrawal of an insurer’s limit can also warn of deterioration before it becomes public. That extends the information available after approval as well as at onboarding. Insurance still leaves the seller with responsibility. Standard policies cover a defined share of the insured receivable, so the seller retains part of the risk and an incentive to set limits carefully.

Building in review triggers so the initial limit does not become permanent by default

An initial limit is based on the little that is known at the time. It should not become permanent through administrative inertia. The policy should define interim review triggers so that new evidence prompts reassessment before the next scheduled meeting. Those triggers can initiate the process automatically while authorized people retain approval responsibility.

Utilization is one useful trigger, with 80% of the limit serving as an example of a defined review point. Reaching it may indicate that the relationship needs more room or that exposure is concentrating. A missed or partial payment needs to be read differently because the buyer is paying in a way that differs from what was assumed at approval. Requests for longer terms and sudden order spikes also warrant review. So do leadership changes, ownership transfers, sector stress, bureau score movements and new liens or judgments.

Scheduled reviews provide a backstop between those events. New accounts need closer monitoring during the first several months, when the buyer’s payments put the opening assessment to the test. High-exposure accounts need quarterly reviews and continuous payment-behavior alerts, while mid-tier accounts need annual reviews and checks on payment patterns. Good performance should earn a route upward. Consistent on-time payment over a defined period supports a higher limit and, where appropriate, longer terms, allowing larger orders and stronger relationships.

The accounts receivable aging report shows whether the framework is working well. If a disproportionate share of receivables is overdue, the weakness may lie in initial credit decisions, collections or both. Approval alone does not collect an invoice. Supplier portals, customer queries and systematic escalation still require persistent follow-up when payment stalls. That operational work connects a defensible credit decision to cash received.

Each review then adds the evidence missing at onboarding. For an established account, payment behavior with the seller is more predictive than an external credit report, making later limits more accurate than the opening estimate. The framework’s value compounds as that record develops. Careful credit management gives buyers who pay reliably room to grow and helps finance spend more time on strategy and less on chasing write-offs it could have avoided.