Hary Periya

1.16

1 //The challenge

I worked with a multi location retailer that was excellent at marketing demand and consistently bad at predicting it.

  • Their best selling items sold out in some stores while sitting unsold in others

  • Overstocked seasonal inventory tied up cash that should have gone toward what was actually moving

  • Store level reordering relied on manager intuition that varied wildly from location to location

2 //The Solution

I helped this retailer build a demand forecasting system that looked at sales velocity, local trends, and seasonality by individual store, not just at the regional level.

The forecast fed directly into reorder recommendations, so the gap between what a store needed and what a manager guessed disappeared.

  • Store level forecasting replaced regional averages that masked real local demand patterns
  • Automated reorder suggestions flagged fast movers before they actually hit zero stock
  • Inventory transfers between nearby stores caught imbalances before markdowns became the only option

Retail does not have a demand problem. It has a visibility problem dressed up as one.

Hary Periya

3 //My Pesonal Thoughts

Every retailer I meet already has the sales data. Almost none of them are using it at the store level.

  • Regional forecasting hides the exact problem it should be solving
  • Store managers were never the problem. They were guessing because nobody gave them better numbers
  • The fix here was not more inventory. It was moving the inventory that already existed to where it was actually needed

4 //Key Outcomes

  • Stockouts on top selling items dropped significantly across the pilot store group
  • Excess seasonal inventory carried into the next quarter fell by a meaningful margin
  • Store managers stopped overriding the system once it proved more accurate than their own estimates
Stockouts Reduced
0 %
Inventory Dollars Freed
$ 0 +