Collection KPIs Every Finance Manager Should Track (and Three to Ignore)
Most collections dashboards measure effort and call it performance. These are the six metrics that actually describe whether receivables are converting to cash, how to calculate each one, and the three popular numbers worth dropping.
On this page
Key takeaways
- DSO alone is not a performance metric. It moves with payment terms and sales seasonality, so it can worsen in a month where the team collected better.
- Pair it with Best Possible DSO — the DSO you would have if nothing were late. The gap between them is the part your team can actually influence.
- Collection Effectiveness Index (CEI) is the closest thing to a true performance measure: it asks what share of collectable money was collected.
- Promise-kept rate and portfolio coverage measure the two things effort metrics only proxy for: whether commitments hold, and whether the book is actually being worked.
- Drop calls made, emails sent and average days to pay in isolation. The first two measure activity, not outcomes; the third hides who is doing it.
Most collections reporting measures how busy the team was. Calls logged, emails sent, accounts touched — numbers that are easy to produce, easy to improve, and almost completely disconnected from whether cash arrived. A team can double its call volume and collect less.
The six metrics below measure outcomes instead. Each one is calculable from data you already have, each answers a question a CFO actually asks, and together they cover the four things that can go wrong in receivables: the terms are too long, the customers pay late, the team is not working the book, or the commitments they extract do not hold.
1. Days Sales Outstanding (DSO)
The headline receivables metric: how many days of sales are sitting unpaid in the ledger.
- Formula: (Accounts receivable ÷ credit sales for the period) × days in the period
- Example: AED 3,000,000 receivables ÷ AED 6,000,000 quarterly credit sales × 90 = 45 days
DSO's value is that it is comparable across periods and across businesses in a way an absolute receivables figure is not. Its weakness is that it conflates two entirely different things: the terms you granted and the lateness of payment against them. A DSO of 60 on 60-day terms is a book being paid perfectly. A DSO of 60 on 30-day terms is a book being paid a month late.
2. Best Possible DSO
This is the fix for the problem above, and it is the metric most teams have never calculated. Best Possible DSO uses only the receivables that are not yet due — it is the DSO you would report if every customer paid exactly on terms and nothing were late.
- Formula: (Current receivables ÷ credit sales for the period) × days in the period
- Example: AED 1,600,000 current ÷ AED 6,000,000 × 90 = 24 days
The number that matters is the gap. With actual DSO at 45 and Best Possible at 24, 24 days are structural — the cost of the terms the business chose to grant — and 21 days are lateness. That 21 is the collections team's addressable target, and it is the only part of DSO that improving the process can move.
3. Collection Effectiveness Index (CEI)
CEI asks the performance question directly: of the money that was available to collect this period, what share was collected?
- Formula: (Opening receivables + credit sales − closing receivables) ÷ (Opening receivables + credit sales − closing *current* receivables) × 100
The numerator is what you collected. The denominator is what you could have collected if everything already due had been paid — which is why closing *current* receivables is subtracted rather than total closing receivables. Not-yet-due invoices were never collectable this period, so holding the team accountable for them would be measuring the calendar.
Read it as a percentage where 100% means everything collectable was collected. Above roughly 80% is a well-run book; below 60% suggests the process, not the customers, is the constraint. Unlike DSO it is unaffected by terms, which makes it the fairest single measure of the collections function itself.
4. Promise-kept rate
Promises kept divided by promises resolved. It is the only metric here that measures customer intent rather than outcome, which makes it leading rather than lagging — it moves before DSO does.
- Formula: promises kept ÷ promises resolved × 100, per customer and per collector
- Exclude unresolved promises from the denominator, or the rate describes nothing
Reported per customer it is a credit signal: a customer who has broken three of their last four commitments is a different risk from one who has kept four of four, regardless of what their financials show. Reported per collector it is a coaching signal — someone logging a high volume of promises that mostly break is accepting soft commitments rather than negotiating firm ones.
It needs at least three resolved promises per customer before the number means anything; below that, one late payment swings it entirely. The mechanics of capturing and resolving promises are covered in payment promise tracking.
5. Portfolio coverage
The metric that answers "is the book actually being worked?" — and the one that most often exposes a gap nobody suspected.
- Formula: overdue value contacted this period ÷ total overdue value × 100
The critical detail is that it is weighted by value, not by account count. A team that contacted 80% of overdue accounts may have contacted 30% of the overdue money, because the small accounts are quicker to work through and the large ones are the difficult conversations. Counting accounts makes that invisible; counting dirhams does not.
This is also the metric that distinguishes a genuinely under-resourced team from a badly prioritised one. Low coverage with high activity means effort is going to the wrong accounts.
6. Aging mix, as a percentage
Not the aging report itself, but each bucket as a share of total receivables — tracked as a trend rather than a snapshot.
| Bucket | June | July | August | Read |
|---|---|---|---|---|
| Current | 58% | 55% | 52% | Falling — overdue population is growing |
| 1–30 | 22% | 21% | 20% | Stable |
| 31–60 | 10% | 12% | 14% | Rising — invoices are not clearing the first bracket |
| 61–90 | 6% | 7% | 8% | Rising |
| 90+ | 4% | 5% | 6% | Rising — recovery exposure building |
Absolute balances can rise with sales volume and tell you nothing. Percentages remove that effect, so the direction is real.
The pattern above is the one to catch early: money accumulating in the middle brackets while current shrinks means invoices are entering the overdue population faster than the team is clearing them. Reported monthly as percentages, it is visible three months before it shows up as a bad-debt provision. How to read the underlying report is covered in the accounts receivable aging report guide.
Three metrics worth dropping
Not because they are meaningless, but because they get reported *instead of* the six above, and they are easy to improve without collecting anything.
| Metric | Why it is reported | Why to drop it |
|---|---|---|
| Calls made / emails sent | Easy to count, feels like productivity | Measures effort, not outcome. Trivially inflated by making easy calls, and it rewards volume over the difficult conversations that actually matter. |
| Number of accounts contacted | Looks like coverage | Unweighted by value, so it flatters a team that worked twenty small accounts and avoided the three large ones. Use value-weighted portfolio coverage instead. |
| Average days to pay, on its own | Intuitive and easy to explain | An average across a mixed book hides everything — a few very late accounts and a compliant majority produce the same figure as a uniformly mediocre book. Useful per customer, misleading in aggregate. |
Putting it on one page
Six metrics is already more than most teams report, and the reporting layer should differ by audience or it will be ignored by all of them.
| Audience | Frequency | Metrics |
|---|---|---|
| Collector | Daily | Their own portfolio coverage, promises due today, promise-kept rate |
| Finance manager | Weekly | Coverage by collector, promises broken, aging mix movement, collections against target |
| CFO / board | Monthly | DSO and Best Possible DSO with the gap, CEI, aging mix trend, concentration |
One discipline matters more than the choice of metrics: compute each one the same way every period, and write the formula down next to the number. Most disputes about collections performance turn out to be disputes about the denominator — and CEI in particular is easy to recompute slightly differently and reach a conclusion the previous method would not support.
Where these get calculated live rather than assembled monthly, they stop being reporting and start being management. That is the argument for holding the aging, the follow-up history and the promise record in one place — see collections management software for the workflow side and accounts receivable software for the ledger it computes from.