A logistics and fulfilment company with a 4,500 m2 warehouse in the Azambuja industrial zone processed over 800 orders per day for e-commerce clients. But the warehouse was organised chaos — operators knew the location of the most popular products by heart, but any new team member took weeks to become productive. The picking error rate was 4.7%, the average order preparation time was 8.3 minutes, and quarterly inventories revealed stock discrepancies exceeding 12%. The implementation of a WMS (Warehouse Management System) with optimised picking and real-time stock reduced errors to 0.4%, picking time to 3.1 minutes and inventory discrepancies to 1.2%.
The Scenario Before: When the Warehouse Depends on Human Memory
LogiStore (real case — data altered under NDA) provides warehousing, picking, packing and dispatch services for 14 e-commerce clients — from fashion and accessories to food supplements and consumer electronics. With 38 warehouse staff and an operation running two shifts (6am–2pm and 2pm–10pm), the company is the invisible link between the online store and the end consumer.
The warehouse was organised organically — that is to say, without real organisation. Products were stored wherever there was space at the time of receipt. The more experienced operators knew where each product was, but that knowledge was personal and non-transferable. When one of those operators was absent or on holiday, shift productivity dropped by 30 to 40%.
The picking process was based on printed paper lists. The operator received a sheet with the items to pick for each order, walked through the warehouse in the order the products appeared on the list (which bore no relation to the physical layout of the warehouse) and placed the items in a trolley. Verification was visual — the operator looked at the product, mentally confirmed it was correct and moved on to the next. No barcode scanning, no electronic validation.
The problems were multiple and interconnected:
Frequent picking errors. Wrong product, wrong size, incorrect quantity, missing item — 4.7% of orders had at least one error. Each error generated a return, a reship and a complaint from the end customer, costing an average of EUR 18.50 per incident.
Inefficient routes. Without route optimisation, operators walked an average of 14 km per shift across the warehouse. Much of that distance was redundant — going to the back of the warehouse for one product, returning to the front for another, and heading back to the rear for a third.
Invisible stock. The exact position of each product was unknown to the system. The ERP inventory stated there were 47 units of product X, but not on which rack, aisle or shelf. Quarterly inventory counts revealed average discrepancies of 12% — products recorded as available but nowhere to be found, and products found that were not in the system.
Slow onboarding. A new operator took 3 to 4 weeks to reach acceptable productivity, simply because they needed to memorise the location of the most frequently picked products.
The Numbers Behind the Problem
• Picking error rate: 4.7% (38 orders with errors per day, out of 800).
• Average cost per error: EUR 18.50 (return + reship + management).
• Average picking time per order: 8.3 minutes.
• Distance walked per operator per shift: 14 km.
• Inventory discrepancies: 12% in quarterly counts.
• Onboarding time for a new operator: 3–4 weeks.
• Maximum daily capacity: 800 orders (operational limit).
• Monthly cost of errors: approximately EUR 21,000.
The Solution: WMS with Intelligent Picking
Component 1: Warehouse Mapping and Addressing
The first step — and the most labour-intensive — was physically mapping the entire warehouse. Each storage location was given a unique address in the format aisle-rack-level-position (for example, A03-E12-N2-P04). Barcode labels were installed at each position. The warehouse was reorganised into zones: zone A for high-rotation products (close to the dispatch area), zone B for medium rotation, zone C for low rotation and zone D for bulky products requiring special handling.
Each product in the system now has one or more associated locations. When new goods arrive at the warehouse, the WMS tells the receiving operator exactly where to store them — considering the product's rotation, available space and proximity to complementary products (products frequently ordered together are stored close by).
Component 2: Optimised Picking with Barcode Scanning
Picking moved from paper lists to handheld terminals (PDAs) with integrated barcode readers. The WMS generates an optimised pick list that minimises the operator's route through the warehouse. Instead of following the list in the order products were ordered, the system organises items by the most efficient travel sequence — aisle by aisle, with no backtracking.
For small orders (1–3 items), the system uses individual picking. For larger volumes, we implemented batch picking — the operator picks items for several orders simultaneously in a single route, placing each item in the corresponding compartment of the trolley. The system automatically calculates which method is most efficient for each batch of orders.
Each item is validated by barcode scan. The operator scans the product barcode and the system confirms whether it is the correct item, in the correct quantity, for the correct order. If the operator picks the wrong product, the terminal emits an immediate audio and visual alert. This validation virtually eliminated picking errors.
Does your warehouse still run on paper lists and operator memory?
We implement WMS with optimised picking that reduces errors and increases capacity without hiring more people.
View Intelligent Automation →Component 3: Real-Time Stock and Continuous Inventory
With the WMS, every stock movement is recorded in real time: receiving (entry into the system with assigned location), picking (exit for order preparation), transfer between locations, return (entry with inspection and reclassification) and adjustment (correction of discrepancies). Stock ceased to be a global number — it became detailed information by location.
We replaced the quarterly inventories (which required shutting down operations for an entire day) with continuous cycle counting. Every day, the system automatically selects a set of locations for counting — prioritising those with the highest rotation or where discrepancies have recently been detected. Operators count those locations during quieter periods, without interrupting normal operations. Over the course of a quarter, the entire warehouse is counted at least once.
The real-time stock dashboard allows the management team and e-commerce clients to see, at any moment, the exact stock of each product, where it is stored, how many units are reserved for orders in preparation and the projected stock-out date based on sales velocity.
The Implementation: A 10-Week Project
Weeks 1–2: Physical warehouse mapping, installation of address labels, WMS configuration and product data import.
Weeks 3–4: Physical warehouse reorganisation by rotation zones. Complete inventory with location recording for each product.
Weeks 5–7: Operator training on PDAs and the new picking process. Parallel operation (legacy system and WMS running simultaneously) for two weeks.
Weeks 8–10: Full transition to the WMS. Configuration adjustments, picking route optimisation and cycle counting implementation.
Results After 6 Months: Before vs. After
• Picking error rate: from 4.7% to 0.4% (91% reduction).
• Average picking time per order: from 8.3 minutes to 3.1 minutes (63% reduction).
• Distance walked per operator per shift: from 14 km to 7.8 km (44% reduction).
• Inventory discrepancies: from 12% to 1.2%.
• Onboarding time for a new operator: from 3–4 weeks to 3 days.
• Daily capacity: from 800 to 1,350 orders (with the same team).
• Monthly cost of errors: from EUR 21,000 to EUR 2,200.
• E-commerce client satisfaction: NPS from 42 to 78.
Financial and Strategic Impact
The total investment — WMS, PDAs (24 units), warehouse labelling, physical reorganisation, integration with client systems and training — was EUR 47,000. The direct saving from picking errors alone totals EUR 225,600 per year. The additional capacity (from 800 to 1,350 orders/day) allowed the company to accept 4 new e-commerce clients without hiring more operators, generating estimated incremental revenue of EUR 380,000 per year. The payback period was under 3 months.
Strategically, the WMS transformed LogiStore's value proposition. The company went from "warehouse with people" to "logistics operation with guaranteed SLAs and real-time KPIs". New clients are attracted precisely by the transparency and reliability the system offers — something that local competitors still cannot match.
Conclusion
A warehouse without a WMS is like a business without accounting — it works until the day it stops working. LogiStore operated for years based on operator memory and paper lists. The system worked, up to a point. But the invisible cost of errors, inefficiency and dependence on key individuals was real and growing. With the WMS, the company not only eliminated those costs but multiplied its capacity and transformed itself into a benchmark logistics partner.
If your warehouse still depends on paper and memory, the potential for improvement is immense — and the return is swift.
