Case 06 Retail · 2025
Replenishment, rebuilt after the spreadsheet.
Store managers spent a day a week ordering stock from a spreadsheet only one analyst understood. We fixed the process, then automated it.
- Client
- A home-goods retailer with 46 stores
- Services
- Digital operations
- Duration
- 16 weeks
- Team
- Iva Stoyanova, Sofia Lindqvist
What changed
Illustrative outcomes for a fictional client.1
9.2%3.1%
Stock-outs on core lines
112
Store hours spent ordering each week
7458
Days of inventory
The story
The situation
Each of the 46 stores ordered stock weekly from a central spreadsheet maintained by one analyst. Managers overrode its suggestions freely, warehouses received orders on different days, and the business carried 74 days of inventory while still running out of bestsellers.
What we found
The spreadsheet’s logic was sound; the process around it was not. Orders arrived on different days, overrides went unrecorded and nobody reviewed forecast accuracy. Buying software at that point would have automated the chaos.
Buying software at that point would have automated the chaos.
What we did
We fixed the ordering calendar and override rules first, measured forecast accuracy for six weeks, and only then specified a replenishment module for the retailer’s existing ERP. We oversaw the vendor through configuration and a rollout in three regions.
What changed
Stock-outs on core lines fell from 9.2% to 3.1%, store ordering time from eleven hours a week to two, and inventory from 74 to 58 days. The analyst now owns forecast quality rather than the spreadsheet.
Timeline
16 weeks
- DiagnoseWeeks 1–3
Ordering process and data flow mapped.
- DesignWeeks 4–6
Calendar, override rules and requirements.
- ImplementWeeks 7–14
ERP module configured, three regional rollouts.
- Hand overWeeks 15–16
Forecast quality review owned in-house.
The team
- Northbridge and this client are fictional. Figures are demo content written to show how outcomes would be reported, not results of a real engagement.