Optimising inventory levels

Knowledge · Inventory optimisation

Optimising inventory means achieving the same service level with less capital tied up. The lever is not blanket cuts across all items, but precision. Deriving safety stock item by item from a reliable demand forecast – instead of from flat coverage rules of thumb – typically lowers capital lock-up by 15–30 % without hurting availability. In principle it takes three steps: forecast demand per item, calculate the required safety stock from forecast spread and target service level, and set the reorder point accordingly. This article explains how, and links the calculators to estimate your own potential.

What does inventory optimisation actually deliver?

Inventory ties up capital and hides process problems, while too little stock is expensive too: shortages, expediting, lost orders. Optimisation means finding, for each item, the point where the sum of capital and shortage costs is minimal – not a gut-feel compromise across the whole range.

The effect is measurable: at 8 % cost of capital, every million euros of inventory freed up equals roughly €80,000 in interest per year, plus warehousing and obsolescence costs. A 15–30 % reduction quickly adds up to a six-figure annual figure.

How does the optimisation work in principle?

The core is separating the expected value from the uncertainty. Base demand over the replenishment lead time is plannable – it follows from the forecast. Safety stock only covers the spread around that expected value. The better the forecast, the smaller the spread to hedge, and the lower the safety stock at the same service level.

Flat coverage ("four weeks for everything") ignores that a steadily selling A item needs far less buffer than a sporadic C item. That is exactly where the savings potential lies: stock is reduced where it is needlessly high and kept only where demand requires it.

Forecast, safety stock, reorder point – the three levers

1. Demand forecast per item: instead of past averages, seasonality, trends and demand patterns are modelled. The result is not just an expected value but also its spread.

2. Calculate safety stock: forecast spread, lead time and target service level yield the item-level buffer. Our safety-stock calculator shows how – and where the textbook formula reaches its limits.

3. Set the reorder point: reorder point = expected demand over the lead time + safety stock. When stock falls below it, a replenishment is triggered – automated, across thousands of items at once.

Your potential

How much capital sits in your warehouse?

Use the capital calculator to estimate how much capital is tied up in your inventory – and what a 15–30 % reduction means for you. For implementation on your own data, we prove the business case in a Proof of Value.

Frequently asked questions about inventory optimisation

How do you calculate the optimal inventory level?

The optimal level is the expected demand over the replenishment lead time plus a safety stock that covers demand variability up to the desired service level: reorder point = average demand × lead time + z × σ × √lead time. What matters is that σ (the spread) comes from an item-level forecast rather than a blanket estimate – otherwise stock ends up systematically too high.

Which inventory level is too high?

Too high is any level that hedges a higher service level than the business requires, or that is based on flat coverage rather than actual demand spread. Warning signs: months of coverage on easily forecastable items, the same buffer for A and C items, and rising stock without rising sales. Comparing flat vs. calculated safety stock per item reveals the excess.

How much can inventory realistically be reduced?

In manufacturers with grown, flat replenishment rules, a 15–30 % reduction at unchanged availability is typical. The exact figure depends on how far today's stock sits from the forecast-based optimum – which can be quantified up front on historical data before going live.

Doesn't cutting stock reduce availability?

No – if done right. The point is to reduce stock where it barely supports the service level (over-buffered, easily forecastable items) and keep it where demand fluctuates. The target service level stays constant or even rises; only the capital deployed goes down.