Every inventory decision is a bet placed with two currencies. One is cash: the working capital tied up in stock that could otherwise fund growth, reduce debt, or simply sit on the balance sheet. The other is service: the probability that a customer finds what they ordered, when they expected it.
Most organisations manage these currencies in separate rooms. Finance pushes for lower days inventory outstanding. Sales and customer service push for higher fill rates. Operations sits in between, absorbing contradictory targets and often settling them through negotiation rather than analysis. The result is an inventory position shaped by whoever argued last, not by a clear view of what the money is buying.
The relationship between working capital and service level is not a matter of opinion. It follows a predictable mathematical shape, and that shape can be mapped, segmented, and translated into financial language. Doing so turns a recurring argument into a structured decision. Three frameworks make this possible: the efficient frontier, segmented investment strategies, and a communication approach built for the CFO's office.
Efficient Frontier Analysis
The starting point is recognising that service level and safety stock do not move in proportion. Under the common assumption of normally distributed demand, safety stock scales with a service factor, or z-score. Moving from a 95% to a 99% cycle service level raises that factor from about 1.65 to 2.33. For a single item, that means roughly 41% more safety stock. Pushing to 99.9% takes the factor to 3.09, adding about another third on top. Each additional point of service costs more than the last.
Calculating this relationship across an entire portfolio of SKUs produces a curve. Total inventory investment sits on one axis and aggregate service level on the other. This is the efficient frontier. Each point on the curve shows the minimum inventory needed to reach a given service level, assuming stock is allocated optimally across items. The idea comes from portfolio theory in finance, and it carries the same discipline. You cannot improve one dimension without paying in the other, unless you are currently operating inefficiently.
That last condition is where the practical value lies. Most companies sit below their frontier, not on it. They hold excess stock of slow, predictable items while running short on volatile ones. Often this happens because safety stock rules were set years ago and never recalibrated. Plotting the current position against the frontier reveals the gap. Closing it is a reallocation exercise that can improve service and release cash at the same time. That combination is rare in supply chain trade-offs.
The frontier itself is not fixed. Several changes shift the whole curve towards less inventory at every service level: shorter lead times, lower lead time variability, better forecast accuracy, and shorter review cycles. This distinction matters strategically. Moving along the curve is a trade-off decision. Moving the curve is an investment in capability. Both are legitimate, but they should be evaluated differently.
TakeawayBefore debating whether to trade cash for service, check whether you are on the frontier at all. The cheapest improvements usually come from fixing misallocation, not from choosing a new point on the curve.
Segmented Investment Strategies
A single service target applied across all products is simple to administer and wasteful in economic terms. It treats a high-margin item ordered by strategic accounts the same as a low-margin item with readily available substitutes. Service gets expensive quickly at the top of the curve. As a result, uniform targets tend to overspend on items where stockouts matter little and underspend where they matter most.
Segmentation starts with the dimensions that determine the value of availability. Classic ABC analysis ranks items by revenue or volume. XYZ analysis adds demand variability, separating stable items from erratic ones. More mature approaches add gross margin, customer criticality, substitutability, and the downstream consequences of a stockout. A stockout might halt a customer's production line or cost a contract. A spare part that idles a mining truck carries a very different stockout cost than a standard fastener, even if their unit values are similar.
Each segment then receives a service target that reflects its marginal economics. High-value, predictable items can often support very high service at modest cost because their variability is low. Low-value, erratic items may justify lower targets or a different policy altogether, such as make-to-order, centralised stocking, or postponement. The aim is not to maximise service for every item. It is to maximise the value delivered by the total inventory budget.
The effect at portfolio level can be substantial. As an illustration, consider a distributor with a blanket 97% target. It might find that differentiated targets, such as 99% for critical items and 90% for the long tail, deliver the same weighted service with materially less stock. Alternatively, they could deliver better service with the same stock. The frontier becomes a working tool here. Each segment has its own curve, and the optimal allocation equalises the marginal return of the last dollar invested across segments.
TakeawayService level is a resource to allocate, not a uniform standard to enforce. The right question is where the next dollar of inventory buys the most customer value.
CFO Communication Frameworks
Inventory conversations often stall because operations and finance use different units. Operations speaks in fill rates, weeks of cover, and safety stock days. Finance speaks in cash, return on capital, and cost of funds. A proposal to raise service from 95% to 98% means little to a CFO in those terms. It needs to be expressed as three things: the extra working capital required, the annual carrying cost of that capital, and the revenue or margin it protects.
Carrying cost is the first translation. It typically includes the cost of capital, storage, insurance, obsolescence, and shrinkage. It is often estimated at 20% to 30% of inventory value per year, though the figure varies widely by industry. An additional $5 million in safety stock at a 25% carrying rate costs $1.25 million a year. That number can then be set directly against the margin it protects. That includes avoided lost orders, expediting costs, and penalty clauses.
The efficient frontier becomes a financial chart once its axes are converted to dollars. Showing the marginal cost of each additional service point makes the diminishing returns explicit. For instance, compare the cost of moving from 96% to 97% with the cost of moving from 98% to 99%. Several metrics link inventory policy to the balance sheet and income statement the CFO is accountable for. These include days inventory outstanding, the cash conversion cycle, and gross margin return on inventory investment.
The framing also matters. Presenting inventory as an investment portfolio, with segments, expected returns, and risk exposure, matches how finance already evaluates capital allocation. It moves the discussion from defending stock levels to choosing among options with known costs and benefits. That shift tends to produce more durable decisions, because both functions have agreed on the logic rather than just the number.
TakeawayA CFO rarely objects to spending working capital. They object to spending it without a visible return, so make the return visible.
The tension between working capital and service level is real, but it is not arbitrary. It follows a curve that can be measured, and that curve reveals both genuine trade-offs and avoidable waste.
The three frameworks build on one another. The efficient frontier shows what is possible and where current policy falls short. Segmentation applies that logic product by product, directing investment where availability creates the most value. Financial translation makes the resulting choices clear to the people who control capital. Inventory stops being a cost to minimise and becomes a portfolio to manage.
The organisations that handle this well are not the ones holding the least inventory. They are the ones that know exactly what each dollar of stock is buying.