Why We Built Our Own Warehouse Software
By Salt of the Earth
The problem no off-the-shelf system could solve
Our warehouse holds 116 active products across five formats and sixteen fragrances, spread over 650 pallet locations, and every pallet carries its own carton size and batch number. A useful system needs to understand all three at once.
Over the years we tried several commercially available warehouse management platforms, including some major ones, and none of them managed it. That left paper, and a paper stock check across 650 pallet locations is not merely a nightmare, it is close to impossible to get right. An inaccurate stock figure quietly spreads into everything else: production planning, purchasing, and what we can confidently promise to customers.
A working prototype in a few days
Ahead of our end of year stock count, our managing director Thomas Laird had an idea for a stock checking app built around the three things the off-the-shelf systems kept tripping over: our carton sizes, our warehouse bin locations and our batch numbers. He built a working prototype in a few days using AI-assisted coding (vibe coding, as the internet insists on calling it), and it worked. People around the company were genuinely impressed, and a little surprised.
"It was a really good proof of concept. It showed us there was potential to massively improve our warehouse efficiency, something we had been chasing for years without success through commercially available software."
The prototype also revealed the next step. Bin locations are only useful if something is actually managing the bins, so the stock taking app needed a proper warehouse management layer on top of it.
From prototype to production
Knowing the idea worked, we brought a professional development team on board to build it properly: stock takes, plus the everyday essentials of warehouse management, picking, packing and booking in. A few months later, the new and improved app has completed its first successful full stock take and has been running picking and packing for the past month.
There was real internal scepticism about whether a deodorant manufacturer could build something like this in-house. The stock check settled it. Version two is already underway.
