Every order arrived twice.
Orders came in as email PDFs. Someone opened each one, typed it into the ERP, then typed it again into accounting. On a good day it took an hour. On a busy Monday it took the whole morning.
The team had tried two off-the-shelf tools. Both solved half the problem and added a third place to check. They did not need another tool. They needed the part between the ones they had.
The process nobody had drawn.
We started with one session and one whiteboard. Four people walked us through an order from the first email to the paid invoice. Nobody had seen the whole thing end to end before.
The map had 23 steps. Eleven of them were someone copying data from one screen to another.
Where the hours went.
We timed each manual step for a week. Three of them accounted for almost all of the time.
Measured over one week, multiplied by 4.3. Illustrative.
“We thought the problem was the ERP. It was the eleven steps around it.”Operations lead
Building the connecting layer.
We kept the ERP and the accounting software. We built one small system that reads each order email, checks it against stock and prices, and writes it to both. Anything unusual goes to a review queue instead of guessing.
The build took six weeks: two for the order intake, two for the accounting sync, two for the customer portal and testing with real orders.
The first Monday without typing.
We ran the new system next to the old process for two weeks. When both produced the same numbers, the team switched. The review queue caught nine orders in the first month, mostly new product codes.
78 hours a month back.
The hours did not disappear. The person who typed orders now calls customers who have not reordered in 60 days. That part is harder to measure, so it is not in the numbers above.
Illustrative example. Staff cost from the team's own estimate.
Map your own order process.
One session, one process, a map you keep. We show you where the manual hours are and what it would cost to remove them.