- A credit policy covers who gets credit, how much, on what terms, and what happens when they don't pay.
- Score customers on payment history, size and financial strength, payment behavior and trend, and exposure.
- Give each tier different limits, terms, review frequency, reminder cadence and escalation triggers; otherwise it's only a label.
- Move accounts between tiers on trigger events, and review riskier tiers more often.
- Tier D (watch) means selling with conditions, not stopping sales.
Most credit policies decide who gets credit, then treat every approved customer the same way. One customer of ours, a contingency-fee employment law firm, ran its receivables with no customer tiers at all: every account got the same calls, the same reminders and the same patience, regardless of how it had paid before. Not every business can choose who owes it money, but every business can choose how it treats each account.
This page gives you a credit policy template you can copy and adapt, plus the part most templates leave blank: how to sort customers into risk tiers, and what limits, terms and collection rules each tier gets.
What a Credit Policy Should Cover
A credit policy is the written set of rules for who gets credit, how much, on what terms, and what happens when they don't pay. It lets sales, finance and collections make the same decision the same way, and it gives you something to point to when a customer asks for an exception.
A complete policy usually covers:
- Purpose and scope
- Roles and approval authority
- New-account requirements
- Customer risk tiers
- Credit limits and payment terms
- Reviews
- Collections and escalation
- Order holds, exceptions and overrides
- Bad debt and write-offs
- Policy ownership
Most templates handle the first three well. Where they stop is differentiation: they say "assess creditworthiness" and then give every approved customer the same terms and the same follow-up. Credit conditions don't have to be identical for every customer, and in practice they shouldn't be. The tiers are what make the rest of the policy usable.
The Credit Policy Template (Copy and Adapt)
Replace everything in {braces}. Keep each section short; the detail lives in the tier matrix further down.
1. Purpose and scope. This policy sets how {Company} extends trade credit to customers of {entities/business units}, invoiced in {currencies}. It applies to all new and existing accounts with payment terms.
2. Roles and approval authority. The {credit manager/controller} owns this policy. Credit limits up to {$X} are approved by {role}; up to {$Y} by {role}; above {$Y} by {CFO}. Sales may request limits and exceptions but does not approve them.
3. New-account requirements. Every new account provides a completed credit application and {number} trade references. Accounts requesting more than {exposure threshold} also provide {financial statements / bank reference}. New accounts start in Tier {C} until they have {6–12} months of payment history with us.
4. Customer risk tiers. Each account is assigned to Tier A, B, C or D (watch) using the scoring criteria in Appendix 1. Finance assigns the tier; sales is informed.
5. Credit limits and payment terms. Limits and terms are set by tier as shown in Appendix 2. No account may exceed its limit without approval under Section 2.
6. Reviews. Tiers and limits are reviewed on a schedule set by tier (Appendix 2) and whenever a trigger event occurs: {list your triggers}.
7. Collections and escalation. Reminder cadence, escalation steps and the owner of each step are set by tier (Appendix 2). Escalation beyond {step} requires approval from {role}.
8. Order holds, exceptions and overrides. New orders are held when an account {is over its limit / has invoices more than X days past due / is in Tier D without approval}. Any override is approved by {role}, recorded with a reason, and reviewed at the next scheduled review.
9. Bad debt and write-offs. Balances are written off when {criteria}, with approval from {role}. Write-offs are recorded and reported {monthly/quarterly}.
10. Ownership and review. Policy owner: {name/role}. Effective date: {date}. Next review: {date}.
Appendix 1 is your scoring sheet, and Appendix 2 is your tier matrix. The next two sections show how to build both.
How to Sort Customers Into Risk Tiers
Bank examiners grade loans as pass, special mention, substandard, doubtful or loss; trade credit can borrow the logic, not the rules. The useful idea is a small number of graded tiers plus a watch state, where each grade changes what you do, not just what you call the account.
Score each customer on four inputs.
Payment history
Your own ledger is the best evidence you have, because it shows how this customer pays you, not how it pays others. Look at average days beyond terms, the share of invoices paid on time over the last 12 months, broken promises to pay and disputes raised. If you track days sales outstanding at customer level, use that too.
Size and financial strength
Revenue, years in business, ownership, and financial statements once exposure passes your threshold. For new accounts, a quick cross-check helps: lenders often use the "5 Cs" (character, capacity, capital, collateral, conditions).
Payment behavior and trend
How the customer pays matters as much as how fast. Warning signs include paying slower than their own usual pattern, partial or round-number payments, paying only after escalation, and going quiet on channels that used to work. External scores can add a view here; for example, D&B's PAYDEX runs from 1 to 100, and D&B treats 80 and above as low risk of late payment.
Exposure and concentration
Open balance against the limit (utilization), the customer's share of your total receivables, and what's in the order pipeline. A reliable customer can still deserve closer attention if they make up a large share of your AR: if a single account would hurt your cash flow badly by paying late, its tier should reflect that exposure, not just its record.
A simple scoring sketch
Give each input 1 to 3 points (3 = strongest) and add them up. These thresholds are an example; set your own.
| Total score (4–12) | Tier |
|---|---|
| 11–12 | A |
| 8–10 | B |
| 6–7 | C |
| 5 or less | D (watch) |
A worked example. Say a distributor has paid on time for most of the past year but slipped twice in the last quarter (payment history: 2). It's a mid-sized, established business that shared financials when it asked for a higher limit (size and strength: 3). Lately it has started making partial payments against larger invoices (behavior and trend: 1). Its balance sits at about two-thirds of its limit and it's one of your ten largest accounts (exposure: 2). Total: 8, so Tier B, but sitting at the bottom edge. That's a useful thing to know: the account doesn't need tighter terms yet, but the partial payments are the signal to watch at its next review.
Two rules keep the score honest. First, a trigger event can move an account to D regardless of its score; a broken promise on a large balance shouldn't wait for the quarterly review. Second, keep the inputs few. Four inputs scored the same way every time beat twelve inputs nobody updates.
Tier Matrix: Limits, Terms and Collection Rules per Tier
This is Appendix 2 of the template: what each tier actually gets. Values are illustrative.
| Tier A | Tier B | Tier C | Tier D (watch) | |
|---|---|---|---|---|
| Limit method | Usage-based, generous buffer | Usage-based | Lower of usage-based or references | Reduced or frozen; new orders on approval |
| Payment terms | Standard (e.g., net 30–45) | Standard | Shorter (e.g., net 15–30) | Prepayment, deposit or shortened terms |
| Limit increases approved by | Credit manager | Credit manager | Controller | CFO |
| Scheduled review | Annually | Semi-annually | Quarterly | Monthly |
| Reminder cadence | Light: before due, then after due | Standard sequence | Earlier and more frequent | Personal contact from the first missed date |
| Escalation trigger | Well past due, or a dispute left open | Past due beyond a set number of days | First broken promise | Any missed payment |
| Order hold | Over limit only | Over limit or seriously past due | Any invoice past a set number of days | Any past-due balance |
| Security | None | None | Optional (e.g., guarantee) | Deposit, guarantee or letter of credit |
Tier D deserves a note. It isn't "stop selling"; it's "sell with conditions". A customer on watch can keep ordering, but on shorter terms or with security, with every new order approved, and with someone in finance talking to them directly. Most accounts should move out of D, up or down, within a few review cycles; one that sits there for a year usually needs a decision.
Setting the limit. Three methods are common. Suppliers using the net-worth method typically set limits around 10% of the customer's net worth, according to D&B. Trade references show what others have extended. A usage-based limit covers what the customer actually buys: expected monthly purchases × (payment terms in days + a buffer) ÷ 30. For example, a customer buying $40,000 a month on net 30, with a 15-day buffer, gets a limit of $40,000 × 45 ÷ 30 = $60,000.
Collections per tier. The matrix sets rules, not scripts. Tier A customers get automated payment reminders and little else, because their payments mostly arrive. Tier C and D accounts get a person early, and their escalation trigger is a missed commitment rather than a day count; when that happens, follow the steps for a broken promise to pay. At the top of the ladder, a demand letter is the formal step before referral.
Moving Customers Between Tiers
A tier is only useful if it's current. Two mechanisms keep it that way.
Event triggers. Move an account as soon as something changes: payments slower than terms for two months running, a limit breach, a broken promise, a drop in an external score, an ownership change, or a large new order. Downgrades can be quick and approved by the credit manager; upgrades should wait for evidence over time and get a second approver.
Scheduled reviews. Review each tier on its own schedule, from monthly for D to annually for A, and lower a limit when a customer shows strain. A scheduled review should be quick: re-score the four inputs, confirm or change the tier, and note the reason. If most reviews end with "no change", that's fine; the point is that someone looked.
Tell the customer when a change affects them. A shorter term or a new deposit requirement lands better as a clear, early conversation than as a surprise on the next invoice.
The catch is that manually applied tags go stale as the book changes. In Daylit, Smart Labels are rule-defined labels that re-evaluate themselves, so a tier stays current without someone re-tagging accounts; here is why AR customer segmentation goes out of date when it's done by hand.
Common Mistakes When Tiering Customers
- Tiers nobody acts on. If Tier C gets the same reminders and terms as Tier A, the tier is a label, not a policy.
- Too many tiers. Seven grades sound precise but blur in practice. Four is usually enough to change behavior.
- Tiering only at onboarding. A customer that was strong two years ago may not be now.
- Undocumented overrides. A sales exception with no reason and no review date quietly becomes the new policy.
- Limits never revisited. A limit set after a good year should be checked after a weak one.
Putting It Together
Copy the template, score customers on four inputs, attach clear treatment to each tier, and move accounts when triggers fire. That turns a credit policy from a document you file into rules your team uses every day.
What should a credit policy include?
Purpose and scope, approval authority, new-account requirements, risk tiers, limits and terms, reviews, collections and escalation, exceptions, bad debt and an owner. The template above covers each in one short section.
How many customer risk tiers should we use?
In many cases, three or four tiers plus a watch state is enough. The test is whether each tier gets different treatment; if two tiers are handled identically, merge them.
How often should credit limits be reviewed?
On a schedule set by tier (more often for riskier tiers) and whenever a trigger event occurs, such as a broken promise or a limit breach.
What's the difference between a credit policy and a collections policy?
A credit policy decides who gets credit and on what terms; a collections policy decides what happens once an invoice is due. Many companies combine them, as the template above does in Sections 7 and 8, because the tier that sets a customer's terms should also set how quickly you follow up.
Who should approve exceptions to the credit policy?
Someone in finance with authority set in the policy, never the person requesting the exception. Record the reason and a review date for every override.



