Retail
Inventory accuracy exceeds standard across 2.1 billion packages
The inventory accuracy of Albert Heijn's stock at Kuehne+Nagel increased from 98.27% in early 2025 to 99.41% in September. This achieved the KPI of 99.40% in an operation that processed 2.1 billion packages in 2024.
1. The client and the challenge
Kuehne+Nagel handles the inbound, handling, storage, and outbound of Albert Heijn inventory. With 2.1 billion parcels processed in 2024 and a reported total deposit value of €5 billion, this operation demands accurate registration and control of goods flows. In early 2025, stock accuracy stood at 98.27%, below the agreed KPI of 99.40%. As project leads, we were tasked with raising inventory accuracy to the desired level and embedding the improved working method into daily operations.
Core pain
How do you keep physical inventory and records aligned across billions of goods movements? When inventory accuracy falls below standard, the need for checks and corrections increases, leaving daily operations without a reliable foundation.
2. Our approach
We combined project management with a Lean approach focused on simplifying work, reducing errors, and improving process control. Work instructions were streamlined, and stock counts were set up based on the four-eyes principle. At the same time, we introduced automation to facilitate data entry and processing, and made storage across distribution centers clearer. By also improving flow, we not only tackled inventory records, but also the daily execution that determines their reliability.
3. The diagnosis
The stock accuracy of 98.27% made it clear that the alignment between physical inventory and the system records was not yet sufficient to reach the 99.40% KPI. Areas for improvement spanned multiple stages of the process: the clarity and adherence to instructions, count verification, data entry and processing, and warehouse organization. Lead time was also a factor, as faster throughput can reduce average inventory levels. The approach therefore had to go beyond merely correcting discrepancies after the fact: it needed to prevent errors and ensure better control across the entire flow of goods.
4. From insight to execution
We translated the areas for improvement into adjustments in receiving, handling, storage, and dispatch. Simpler work instructions made the desired execution clearer and easier to follow. During inventory counts, the four-eyes principle provided an extra check, while implemented automation layers supported data entry and processing. Inventory was stored more clearly per distribution center. Lead times have been reduced, which led to a decrease in average stock levels. In this way, process agreements, checks, and system support came together in a single, improved way of working.
5. The result
Inventory accuracy rose from 98.27% in early 2025 to 99.41% in September 2025, exceeding the agreed KPI of 99.40%. The 1.14 percentage-point improvement was driven by simplified work instructions, additional checks on counts, automation, and a better-organized storage system. In addition, shorter lead times resulted in a lower average inventory. No separate metrics were provided for this inventory reduction and its financial impact; therefore, no savings percentages or euro amounts are tied to it.
99.41%
Stock accuracy in September 2025 · early 2025: 98.27%
99.40%
Agreed KPI · exceeded in September 2025
2.1 billion
Packages processed in 2024
The volumes relate to 2024; the baseline measurement and the achieved accuracy level to 2025. The increase from 98.27% to 99.41% represents an improvement of 1.14 percentage points. The stated €5 billion refers to the total deposit value and is not a realized saving. A further increase to, for example, 99.9% is a possible ambition, not a proven outcome of the technologies mentioned below.
6. Next step
The next step is to safeguard the achieved accuracy and explore which technologies can support further improvement. AI-driven demand forecasting can help align inventory levels with seasons, promotions, and trends. Where technically and economically feasible, RFID and IoT can enhance the visibility of pallets and load carriers while reducing manual scans and counts. Computer vision for inbound and outbound flows can verify quantities and flag discrepancies, complementing the four-eyes principle. Using a digital twin of warehouse operations, changes to layout and processes can be tested in advance for throughput time and accuracy. AI-powered data analytics can detect recurring patterns in inventory discrepancies, after which the root causes are investigated on the floor. These are potential next steps, not yet realized results: for each application, data quality, feasibility, investment, and return must first be justified and tested in a pilot.