Q1: Which credit KPIs should I actually be tracking?

Four, and they each answer a different question. Read together they are very hard to fake; read individually, each one can be made to look good while something else deteriorates.

The four numbers:

• DSO (Days Sales Outstanding) — speed. How long it takes to turn an invoice into cash.
• CEI (Collection Effectiveness Index) — quality. Of everything you could have collected, how much you actually did.
• Overdue percentage — risk. How much of the book has slipped past terms, and how far past.
• Collection ratio — effort. What the collections team actually pulled in this month.

Q2: My DSO is above the industry average. Is that a problem?

Not necessarily. A benchmark is a question, not a verdict. It tells you where to look; it does not tell you that something is wrong. If your number sits outside the published range, the next step is to find out what the difference is made of.

Two businesses can both sit twenty days above the average for entirely different reasons. One has three large customers on negotiated extended terms, which is a deliberate commercial decision. The other is issuing invoices four days late every month, which is a process failure. The benchmark cannot distinguish between them. Only you can.

Q3: What is the most useful comparison I can make?

Your own number from last quarter. It is the fairest benchmark available to you because everything else is held constant — same customers, same sector, same terms, same team. Direction of travel matters more than position on a table.

In order of usefulness:

• Your number this quarter against your number last quarter.
• Your DSO against your own stated payment terms.
• Your number against a published sector range.

Q4: What is a normal DSO for my sector?

It varies enormously — roughly six-fold from the fastest sectors to the slowest — and almost none of that variation is about how hard the collections team works. It is about the business model. Broad cross-sector medians sit around 40 to 45 days, with the global cross-sector average closer to 59 days.

Q5: How do I tell whether my DSO is too high?

Measure the gap between your DSO and your own stated payment terms, not against a sector table. A DSO of 55 days is tight on 60-day terms and poor on 30-day terms. Same number, opposite conclusion.

Read the band you fall into:

• Within 5 days of terms — your process is working. Document why, because that knowledge leaves when people resign.
• 5 to 15 days over terms — collections drag, and almost entirely fixable in house. Usually late invoicing, invoice errors, or nobody calling before the due date.
• More than 15 days over terms — something structural. Either your stated terms are fiction, you have taken on credit risk you should not have, or a chunk of the book is sitting in dispute.

Convert the gap into money before you present it. On R24 million of annual credit sales, a 33-day gap is roughly R2.2 million of your working capital permanently funding your customers' operations. That figure moves a management conversation in a way that "63 days" does not.

Q6: What counts as a good CEI?

Above 90% is best in class. Published cross-industry averages put the middle of the pack between 70% and 85%.

• 90% and above — best in class. Your remaining risk has moved upstream into credit granting.
• 85 to 89% — strong. A mature process; the gap is usually a handful of named accounts.
• 70 to 84% — where most B2B teams sit. Be careful reading this as a pass mark: at 80%, one rand in every five you could have collected, you did not.
• Below 70% — investigate. Roughly three in every ten collectable rand went uncollected, which is a process problem rather than a market condition.

CEI is more reliable than DSO for judging a collections team, because DSO is distorted by sales volume. A strong sales month can pull DSO down while nothing about collections has changed. CEI only measures what was collectable against what was collected.

Q7: What should my ageing profile look like?

The shape matters more than the exact percentages. Most of the book should be current, and it should fall away sharply as you move right. If the profile bulges in the middle instead of tapering, your collections process is starting too late.

Two invoices of R200 000 are not the same invoice. One at 40 days is an admin task. One at 120 days is a provision waiting to happen.

Q8: Who should own each ageing bucket?

A named person, not a department — and the owner should change as the bucket ages. If the same person owns day one and day 120, nothing escalates, and things that do not escalate get written off.

A workable escalation ladder:

• Current and 1 to 30 days — the credit controller, working proactively before due date.
• 31 to 60 days — the credit manager.
• 61 to 90 days — credit manager with sales in the room, because stop supply is on the table.
• 90+ days — financial manager or CFO, for legal handover or provisioning.

Q9: What are collection ratio and promise-to-pay kept rate, and is there a benchmark?

There is no credible published benchmark for either, and your own first three months of data becomes the benchmark. That is the point rather than a weakness — both measure your process against itself.

Collection ratio is cash collected divided by the overdue balance you opened the month with. Because it resets monthly it answers "was last month a good month", which DSO and CEI cannot do — both are smoothed over time.

Promise-to-pay kept rate is promises kept divided by promises made. Most teams record the promise; almost nobody records whether it was kept, which is where the value sits.

How to start tracking P2P kept rate:

• When you call, get a specific date, a specific amount and the name of the person releasing payment.
• Add one column to your spreadsheet: kept, or not kept.
• Review per customer after two months. A customer who breaks three promises in a row has told you something their financial statements will not reveal for another eighteen months.

Q10: My numbers sit outside the ranges. Where do I start?

Ask one question of every number that sits outside its range: is this my business model, or is this my process? One of those is fixed and the other is cash you can still recover, and almost nobody separates them — which is why the same figure gets reported for three years running.

If it is the business model — long project cycles, retention, public sector payment runs, negotiated terms with your largest customers — your number is a description rather than a problem. Write down the explanation and put it in your board pack before someone without that context draws their own conclusion.

If it is the process — late invoicing, missing PO numbers, unallocated cash, nobody calling before due date, no named owner per bucket — it is recoverable cash, and most of it requires no difficult conversation with a customer at all.

Five action points:

• Pick one KPI, not four, and measure it weekly for a month. One number measured weekly changes behaviour; four measured occasionally change nothing.
• Benchmark against your own terms before any sector table.
• Give every ageing bucket a named owner, escalating as the bucket ages.
• Reconcile before you chase.
• Start recording promise-to-pay outcomes: kept, or not kept.

Need more help?

The support team is ready to assist

support@tradeshield.ai

+27 (0) 10 753 1630

www.tradeshield.ai