All field notes

Amazon · Inventory

Eighty-five to ninety-two

I owned in-stock for the Southwest and took it from 85% to over 92%.

  • Stage: 1 to 2
  • Inventory and planning
85% to 92%+
Regional in-stock
5 to 10%
How close forecasts landed to plan
near 100%
In stock and sold through, Valentine’s Day 2018

The problem

When a customer wants something in an hour, out of stock means lost. I owned in-stock for the Southwest, where the rate sat at 85%, and there was no shared tool for planning it.

What I did

I built a planning tool in SQL and Excel. It set safety stock by rule, adjusted the forecast by region, and planned inventory before and after big events. I wrote a playbook for getting ready for those events, and a way to reconcile finance’s top-down forecast with the item-level one built from the bottom up.

What changed

Regional in-stock went from 85% to over 92%. Forecasts landed within 5 to 10% of plan. On Valentine’s Day 2018 we were close to 100% in stock and sold nearly all of it. Amazon’s economists later scaled the tool, and the playbook was used across the network.

Stuck somewhere between order and cash?

Bring the messiest part. That’s usually where I’d start anyway.

I’ll tell you plainly if I’m the wrong person for it.