Loading
Loading
Before: Morning production started at 4 a.m. off a sheet the lead cook adjusted based on "how Friday felt." Retail managers texted changes that never made it back to the central team until someone noticed empty hooks. Good weeks looked lucky. Bad weeks looked like theft. Nobody had a chart that said "we made 200, sold 140, reason: rain." What was broken: Pars were legacy from when the chain had half the doors. New items borrowed another SKU's par because the system had no history yet. Waste was embarrassing, so it got buried in "samples" or "damaged" without photos. That made the next forecast lie, which made the next overproduction worse.
We pulled twelve weeks of movement by SKU and day. Not pretty charts for investors. A grid the production lead could read at a glance. Production caps scaled with shelf life. High-risk items got smaller batches and a second pull time instead of one big morning push. Toss required a tap on a tablet: reason, photo for high-value. Took ten seconds. Cut the fiction.
Shrink stopped being a moral lecture and became a line item with a name. Sales did not drop when production tightened. They often went up because the right SKUs were actually on the shelf when people showed up. The commissary lead finally slept through the night without guessing.