Working Capital as a Back Office KPI

Ask a shared services manager what they are measured on and you will hear about invoices processed per full-time equivalent, cost per transaction, cycle time and error rate. Ask the treasurer what keeps them awake and you will hear about the cash conversion cycle. These two conversations happen in the same building, about the same work, and almost never reference each other. That separation is the reason working capital KPI programmes usually fail. The cash conversion cycle is not a finance outcome that operations influences indirectly. It is the direct arithmetic consequence of decisions made in order entry, credit control, invoicing, procurement and warehousing, mostly by people who have never been shown the number their choices produce.

Most of DSO happens before the customer sees the invoice

Days sales outstanding is reported as a single figure and treated as a measure of how promptly customers pay. Decompose it and that turns out to be only part of the story. The first component is the lag between delivering the service or goods and raising the invoice. In project businesses and services firms this is routinely a week or more, and it is entirely self-inflicted. The second is invoice accuracy. A wrong purchase order reference, a missing delivery note number, the wrong entity name or an unapproved rate does not generate a phone call; it generates silence, and the clock does not start. The third is delivery: whether the invoice reached the person who can process it, in the format their system requires, through their portal, with the documents attached. The fourth is dispute resolution, which is usually unowned and therefore unbounded. Only the fifth is the customer's actual payment behaviour. In most organisations that measure these separately for the first time, the majority of the addressable delay sits in the first four. That is good news, because those are process problems with owners, whereas customer payment behaviour is a negotiation.

Payables is where measurement goes wrong

Days payable outstanding is the easiest number in the set to improve and the easiest to improve dishonestly. Paying suppliers later extends the metric and transfers the financing cost to whoever has the weakest balance sheet in the chain, which in practice means your smaller suppliers. They recover it through price, or they fail, and either outcome arrives back on your income statement eventually. The legitimate version of payables improvement is different: negotiate terms explicitly rather than taking them by default, capture early payment discounts where the implied return beats your cost of capital, eliminate the duplicate and early payments that most accounts payable functions are quietly making, and use supply chain finance where the arrangement genuinely benefits both sides rather than as a relabelled stretch. The diagnostic question is simple. Are your payment terms the result of an agreement, or the result of your process being slow. A large share of apparent DPO performance is just approval bottlenecks, and it comes with supplier relationship damage that nobody has costed.

Inventory days are a forecasting metric wearing a warehouse uniform

Inventory sits on the balance sheet, so it is treated as a supply chain problem. It is mostly a consequence of two things: how good the demand signal is, and how much the organisation fears a stockout. Safety stock levels in most mid-market businesses were set years ago by somebody reacting to a shortage, and have never been recalculated against actual demand variability and lead time variability. That is why inventory reduction programmes that focus on warehouse discipline produce so little, while the ones that revisit stocking parameters produce a great deal.

Measure it in a way that cannot be argued with

Days beyond terms, not just DSO. A business selling on ninety-day terms and one selling on thirty are not comparable on DSO. Days beyond terms measures what you actually control. Aging distribution, not averages. A mean hides the concentration. Usually a small number of accounts and a small number of disputes contain most of the overdue balance. Dispute rate and dispute resolution time. These two numbers predict next quarter's DSO better than this quarter's DSO does. First-time-right invoicing. The percentage of invoices that require no correction. It is the single most useful operational metric in the whole set and almost nobody tracks it.

Practical Guidance for a Working Capital Assessment

• Decompose DSO into its five components and measure each. Until you know where the days actually accumulate, every improvement initiative is a guess.
• Track first-time-right invoicing weekly. Publish it by team. Invoice quality responds to visibility faster than to any system change.
• Give disputes an owner and a service level. Unowned disputes are the largest single pocket of unnecessary receivable in most businesses.
• Measure days beyond terms alongside DSO. It separates your performance from your customer mix and your sector's conventions.
• Recalculate safety stock against actual variability. Not against the number someone set after a shortage in a year nobody remembers.
• Stop counting late payment as payables performance. Report negotiated terms and actual payment separately so the difference is visible to whoever is being congratulated.
• Put cash metrics in operational scorecards. The order entry team should see the DSO consequence of a missing purchase order reference. That connection is the entire programme.
• Review the largest twenty customer accounts individually every month. Concentration means the tail rarely matters, and a conversation resolves more than a dunning letter.

The Regional Angle

Working capital in the Gulf has structural features that imported playbooks do not anticipate, and the biggest of them is retention. Contracting, engineering and fit-out businesses routinely have five to ten per cent of contract value withheld, released only after a defect liability period that can run a year or more beyond completion. That money is not late and it is not disputed; it is contractually deferred. It distorts every receivables metric it touches, and it should be reported as a separate line rather than allowed to sit inside DSO making the collections team look incompetent. Payment chains here are also longer and more conditional than in most markets. Pay-when-paid arrangements in construction pass the delay down the tier structure, so a subcontractor's cash position is a function of a developer's payment behaviour two levels up. Government and quasi-government customers pay reliably but on their own cycle, and for many businesses those are the largest accounts. Neither of these is a collections problem, and treating them as one wastes the collections function's time on the accounts where nothing can be done. Post-dated cheques remain a working instrument across much of the regional mid-market, and they change the shape of the problem rather than solving it. A cheque dated ninety days forward is a payment commitment with legal weight behind it, which is why it persists, but it is not cash, it can be returned, and organisations holding a drawer of them frequently have a rosier view of their receivables than the position warrants. The change that has materially raised the cost of slow collection is now in its second year. Under value added tax, output tax is generally due by reference to the invoice rather than to payment, which means a business extending long credit is funding the government's share of an invoice it has not been paid for. That was a theoretical point in 2018 and is a visible cash line in 2019. Any regional working capital assessment should quantify it explicitly, because it changes the return on collections improvement considerably and it is the argument most likely to get finance leadership's attention. Two further local factors. Receivables financing and credit insurance are available here but less deep and more expensive than in Europe or North America, so the option of financing your way out of a working capital problem is narrower. And corporate credit information, while improving as regional bureau coverage expands, is still thinner than most credit policies assume, which puts more weight on your own payment history data than on external scoring. Finally, the calendar: approval cycles slow noticeably during Ramadan and the summer months, and a collections plan that does not anticipate that will simply record two bad quarters a year.

The objection worth taking seriously

The objection is that working capital metrics are among the most gamed numbers in corporate finance. Quarter end arrives, payments are held back for a week, shipments are pushed out, collections are chased with unusual enthusiasm, and the reported cycle improves without anything changing. The following quarter it reverses. Anyone who has watched this happen a few times is entitled to scepticism about a programme built on these measures. The harder version of the objection is more substantive. Cash conversion is largely determined by things operations does not control: the sector, the customer mix, the competitive position, and whether you are the small party in a relationship with a much larger one. A supplier to major retailers or to government does not negotiate terms; it accepts them. Meanwhile, squeezing payables to improve the number pushes financing cost onto smaller firms who fund it at higher rates than you would, which makes the supply chain as a whole more expensive and more fragile. Optimising your own cycle at the expense of the chain is value transfer, not value creation. Both points are correct and both are handled by measuring the right things. Days beyond terms, dispute rate, first-time-right invoicing and invoice lag are not sector artefacts and cannot be improved by holding a payment run for a week. They are a description of how well the organisation does the work, they respond to process change rather than to quarter-end theatre, and they improve the supplier and customer relationship rather than mining it. Use them as the scorecard, report the cash conversion cycle as the outcome, and the gaming problem largely takes care of itself.

Common Questions

Who should own the cash conversion cycle?

Treasury owns the reported outcome; the operational components need individual owners in the functions that create them. Invoice lag belongs to the people who raise invoices, dispute resolution to whoever can actually settle a dispute, and stocking parameters to planning. A single owner for the whole number produces reporting rather than improvement.

How quickly can DSO be improved?

Invoice accuracy and invoice lag respond within a quarter because they are internal. Customer payment behaviour takes considerably longer and is partly a commercial negotiation. Expect early gains from the self-inflicted portion and be honest with the board about which is which.

Is stretching supplier payments ever legitimate?

Where terms are agreed openly and priced into the commercial relationship, yes. Where they are achieved by paying later than agreed, no, and the practice tends to return as higher prices, reduced service or supplier failure in the part of your base least able to absorb it.

What should we expect over the next twelve months?

Expect the tax treatment of receivables to keep pushing collection discipline up the regional finance agenda as the second full year of value added tax closes. Expect supply chain finance and receivables programmes to keep expanding in the Gulf as banks build the products. Expect the gradual move towards structured and electronic invoicing across the region to reduce the delivery and format component of DSO, slowly. And expect credit information coverage for regional corporates to keep improving, which will make credit limits a decision rather than a guess.

---

Working Capital Assessment — we split your cash cycle into the parts you cause and the parts you inherit, put numbers on both, and hand the controllable ones to the teams that can actually move them.