Kitting a 600-store rollout without a spreadsheet fire
How we structure BOMs, count sheets and cartons so a store manager can install without calling anyone.
The hard part of a large rollout is never building the fixtures — it is making sure store 412 gets the right count of the right parts with instructions a store manager can follow without calling anyone. That is a kitting and documentation problem more than a manufacturing one.
Six hundred stores means six hundred chances for something to arrive short, mislabeled or without the count sheet matching what is actually in the box. Here is the structure that keeps that from happening.
One BOM, one source of truth
Every SKU in the program lives in a single bill of materials tied to a destination list, not a spreadsheet somebody is manually updating for each store. When a store count changes, it changes in one place and the kitting plan updates from it — not from a second spreadsheet that drifts out of sync with the first.
Count sheets travel with the carton, not separately
A packing list emailed to a regional manager gets lost. A count sheet taped inside the carton, itemized against what is actually packed, gets read by whoever opens the box — usually a store employee with five minutes and no context.
Build the instruction card for the worst case, not the best case
Assume the person installing it has never seen the fixture before and is not going to call anyone. Five steps, no tool references, pictures over paragraphs. If it needs a phone call to install, it needs a redesign, not better instructions.
Six hundred stores is six hundred chances for something to go wrong. The structure has to assume that, not hope around it.
Keep reading
All articles →
Design your display for the freight bill, not the photo shoot
Two cartons per pallet layer instead of one is usually a 20% freight swing. Where to find that in the geometry.
Why we build a first article before a production run
It adds three days and removes the argument. What goes in a first-article report and how to read one.
Tolerance callouts that actually change the price
Half the tolerances we receive cost money without doing anything. A short guide to which features deserve a tight number.