The controller spent the first week of every month rebuilding the same five reports. Sales came from one spreadsheet, purchases from another, and inventory from a third that the warehouse manager updated “whenever he remembered.” At month-end, the sales manager’s numbers and the controller’s numbers differed by about 11 percent, and the owner made decisions based on whichever document was newest. That was the real operating system of a company doing $22 million a year.
When good growth meets bad data, someone always pays
The tipping point came when the company landed its biggest customer ever — a national retailer that ordered $1.4 million in the first six months. The order was great news right up until the fulfillment chaos started. The company oversold two items it didn’t have, underordered one it had plenty of, and shipped three customer orders to wrong warehouses because the address list lived in an email thread.
The customer’s logistics manager called the owner and asked a simple question: “Do you actually know what you have?” He could not honestly say yes. That’s when he called in ERP Consulting.
The review took three weeks and produced a strange document: a map of every spreadsheet, every “final version” file, and every manual transfer between the two. There were 14 active trackers, at least five of which disagreed with each other on basic facts like current stock and open purchase orders.
What the first 90 days looked like
The ERP Consulting team did not replace everything at once. They started with one core: a single shared system for purchase orders, inventory, and sales orders, so that all three functions read from the same set of numbers. The old spreadsheets were frozen on a shared drive for reference but banned from daily decisions.
Conversion of the messy files took six weekends, mostly because the data in the trackers didn’t match and every conflict had to be resolved with a phone call. The team worked through the cleanup in three stages:
- Freeze every spreadsheet and assign one person as the single keeper of each type of data — one for stock, one for orders, one for prices.
- Enter all open purchase orders and open sales orders into the new system before touching historical numbers, so the live picture is complete and current.
- Run a weekly “one version of the truth” check where the warehouse count, the system count, and the spreadsheet count are compared side by side until they agree.
The first clean month-end came 60 days after go-live. The controller closed the books in four days instead of nine, and the sales and finance numbers matched to the dollar for the first time in the company’s history.
The results that show up in the bank account
Three months in, the company stopped paying for 17 percent more inventory than it sold, because purchasing finally saw real stock levels before ordering. The national retailer’s fulfillment complaints dropped from four per week to zero in the second quarter, and the account grew. The owner said the system paid for itself in the first inventory cycle just through fewer rush orders and fewer emergency freight bills.
Your test could not be simpler. Print the current stock list, the open purchase orders, and the open sales orders from your three most important “systems.” If the quantities don’t tell one consistent story, you are running on opinions, not data.
ERP Consulting can help you build the single version of the truth. But before you call anyone, count how many spreadsheets your team updates manually every week — that number is your starting point.