A home goods retailer with three warehouses was bleeding 18% inventory accuracy. Every month, the stock counts in the system disagreed with what was physically on the shelves, and the team was spending 30 hours a week hunting for things that weren’t where they said they’d be. The owner’s favorite line was “the system is wrong” and, honestly, he was half right: the system repeated errors that the picking and receiving teams made every single day.
The turning point happened when a rush order for a wedding in two days couldn’t be fulfilled, because the system said a certain lamp shade was in stock, but the shelf location was empty. The customer left, the manager apologized, and the owner finally agreed that the problem was not the tools but the process. ERPNext Consulting started by counting the fastest-moving 100 items across the three sites, and just that exercise exposed a shocking rule: the most popular items also had the most inconsistent storage locations.
The three fixes that took the accuracy from 82% to 97%
- First, every shelf gets a permanent address. The team stopped moving products to “temporary” spots after receiving, and any item found outside its shelf is treated as a mini-incident that gets logged with a reason.
- Second, daily counts of the top 50 sellers per warehouse, done by the same person, at the same time, using a consistent counting method. Ten minutes a day, not a huge count at year end.
- Third, transfer orders between warehouses must be posted the same day, not at the end of the month. This single rule eliminated most of the phantom inventory across all three sites.
By day 90, the accuracy had risen to 97%, and the emergency exception count dropped from dozens a week to three in the whole month. The picking team stopped losing time, the customer service team stopped promising items that did not exist, and the owner stopped confusing “the system is wrong” with “the workflow is wrong”. ERPNext Consulting came in as a guide, but the real ownership of the change landed with the warehouse supervisors who adopted the shelf addresses as their own idea.
What most warehouse managers still get wrong
They buy nicer equipment, more labels, more screens, and assume the accuracy improves automatically. It doesn’t. The accuracy improves when every move has a consequence: a shelf gets a location, a transfer gets posted the same day, a miscount gets explained. That’s a discipline problem, not a technology problem. Once the discipline is in place, the technology actually starts to work, because it’s fed with accurate information at every step.
For a business with multiple warehouses, don’t start with a full inventory count. Start by counting the 100 fastest-moving items, because those are the ones doing the most damage when they’re wrong. Track how long it takes you to reconcile those 100 items and why the errors occur. Fix the top three causes, then expand the routine. That’s how you get from “the system is wrong” to a number you can actually trust.