What a Logistics Director Is Really Measured On: The Holy Trinity of KPIs

In this episode of Rozmowy Logistyków, Piotr Skobało and Adam Sobolewski discuss what a logistics leader is measured on — KPIs in the supply chain. Piotr talks about what he was personally accountable for while building the supply chain at Cersanit, and together they cover what a COO should be discussing with the board.

COST TO REVENUE: LOOK AT THE PERCENTAGE AND THE ABSOLUTE VALUE TOGETHER

Logistics costs relative to revenue look completely different depending on the business model: retail chains usually sit at 2-4%, while e-commerce with a fragmented last mile (shipping piece by piece, not full pallets) can reach the low double digits. Key protip: look at the percentage and the absolute value at the same time — 1% in an organization with revenue in the hundreds of millions is still a huge amount worth analyzing, even if it looks insignificant as a percentage. In very large organizations (e.g. Amazon), accounting goes down to "bips" — hundredths of a percentage point — where a 5-6 bip improvement is already a major win.

SERVICE: ON-TIME IS ONLY THE START

Service level (On Time in Full) measures both the availability of goods at the point of order fulfillment and delivery within the promised time — and each of these breaks down further into dozens of detailed root causes (was the truck late, did production not keep up, did the warehouse plan the loading badly). In retail, availability is sometimes measured in layers: availability against the minimum shelf-display quantity (there must be at least 3 units for the shelf to look right), availability against the sales forecast between deliveries, and aggregated availability by region and category. A trap to avoid: measuring service against a delivery date the customer changed rather than the original one — one pharmaceutical company kept overwriting the "promise date" after every change in customer preference, artificially inflating the metric without any real improvement in operations. Good practice is to track both dates in parallel — the original one (for root-cause analysis) and the updated one (for holding the team accountable, when the change came from legitimate customer communication).

INVENTORY TURNS: AVAILABILITY, NOT AN ARBITRARY DAYS-OF-COVER TARGET

Inventory turns (in manufacturing companies) and Stock Cover (in retail) versus stock aging are two different things, often confused. The key mistake: setting a rigid target like "20-30 days of cover" instead of treating Stock Cover as an outcome derived from product availability for sale. A real example: cutting Stock Cover by 40% over three years while simultaneously raising availability by 2 percentage points to 99%+ — possible precisely because the goal was service, and inventory turns came out of that as a side effect, not the other way around. Working capital acts as a natural brake here on the instinct to "buy more so we always have availability" — because every extra unit of stock is frozen cash. Aging (goods sitting for more than 90/180/365 days) is a separate measure — goods sitting for over a year require serious reflection, even without a formal expiry date, because they lose their appeal against new collections and trends.

SAFETY AND QUALITY AS THE FOURTH PILLAR

Beyond cost, service, and inventory turns (the "holy trinity" of logistics), a fourth group is quality and safety — metrics like picking accuracy or damage rates, measured both in quantity and value (damaging 1 unit in 1,000 sounds like a success at 99.9% accuracy, until that one unit cost tens of thousands of złoty). Operational efficiency is a function of cost relative to quality — improving one metric without controlling the other is a zero-sum game: full-pallet shipping lowers transport cost but raises working capital and Stock Cover, because there's less flexibility to order and sell smaller units.