Credit and Collections KPIs That Matter

A credit function without metrics turns into a battle of opinions: sales thinks credit is blocking deals, credit thinks sales is bringing bad accounts, and nobody has a number to back their position. The metrics below are the ones that actually change decisions — and there are not many.

Delinquency rate

Delinquency (%) = balance past due beyond X days ÷ total receivables

The cutoff has to be defined and stable: 30, 60 or 90 days. Comparing delinquency measured at different cutoffs means nothing. Track it by segment and by cohort as well — the overall average hides the information that matters.

Aging

Distribution of receivables by days past due:

Aging is the day-to-day workhorse because it shows where the money is and what to do with each block.

Days sales outstanding (DSO)

DSO = (accounts receivable ÷ credit sales for the period) × days in period

It measures how long, on average, it takes to collect. Its value is the comparison against the terms your policy grants: a DSO well above average terms means the agreement is not being honored — and working capital is financing the customer.

Roll rate

The share of the book that moves from one aging bucket to the next month over month.

Roll rate = balance that moved from 30 to 60 days ÷ balance that was at 30 days

It is the best leading indicator in collections: when roll rate rises, the next quarter's delinquency is already booked. It lets you act before the bad number appears.

Recovery rate

Recovery (%) = amount collected in the period ÷ eligible past-due balance

Measure it by aging bucket and by collection channel. Recovery falls fast with age: what you collect within 30 days is an order of magnitude above what you collect after 180. That single fact is what justifies investing in preventive collections.

Underwriting metrics

• Approval rate — applications approved ÷ applications reviewed
• Average decision time — by dollar band and by path (automated, analyst, committee)
• Automated decision rate — share resolved with no human
• Exception rate — share approved outside policy

Exception rate deserves special attention. Above 10%, the policy no longer describes the operation.

Reserves versus actual losses

The allowance estimates future loss; write-offs record what became a loss. Tracking both in parallel shows whether the reserve is calibrated — a reserve consistently above realized loss distorts reported results as much as an insufficient one. The calculation is in allowance for doubtful accounts.

A one-screen dashboard

1. Total delinquency and aging
2. Roll rate for the month
3. DSO against average terms granted
4. Concentration in the top 10 accounts
5. Approval rate and average decision time
6. Exception rate
7. Recovery by bucket

What to take from this

Pick a few metrics and keep the methodology stable. Aging and roll rate drive the month; cohort performance, exception rate and concentration expose structural problems. And always read credit next to sales — the isolated number misleads.