Series: AI × Supply Chain build breakdown
Stack: Claude Code · Google Sheets · co-manufacturer stock counts
The problem
It’s easy to plan how many units to make and forget you’re out of the pouch they go in. Packaging has long lead times, big minimums, and it sits at the co-man, off to the side of the main inventory view. A pouch shortfall stalls a run just as hard as running out of product, it just blindsides you instead.
What I built
Packaging into the plan: pouches, film and boxes tracked per SKU against the production schedule, each with its own lead time and order point. A run about to be blocked on packaging now flags weeks out, not the morning of pickup.
Where it bit me
Two things, and the second one stung. One: the minute packaging was in the plan, it showed that a single flavour’s pouch, not the product, was the real limit on the next run. The thing capping growth was something nobody was even looking at. Two: the physical count came back well under what my books said. A yield or booking gap somewhere upstream.
Lesson: a planner is only as honest as the count behind it.
Leverage rating: 4 / 10
Recurring, and it catches a whole class of stockout the product plan can’t see.
Related: this logic lives inside the Supply Planning Agent, which plans packaging alongside product. Email me to compare notes.