Hundreds of products forecast across
75 distribution centers

Case study

A purchasing cooperative for a global
 food chain

Industry
Food and
beverage
Users
Demand planning, procurement, category management
Scope
75 DCs, demand only
The Challenge
75 DCs feeding restaurants that expect full deliveries on schedule.
Scale beyond manual reach - so forecasting ran at a high level, not at the granularity decisions needed.
Only the top items got real attention. The rest were managed by buffer and judgement.
Highly sensitive to point-of-sale movement and promotions, both of which move fast.
A control tower under construction, but no demand view to purchase against.
What Hexight Did
Demand module sits on top of the existing stack — nothing replaced.
More than 14,000 products forecasted weekly across 75 DCs, autonomously.
A volume beyond what any team could review item by item.
Tactically: better allocation of product across the DCs.
Strategically: commodity hedging on a demand view they
can rely on.
Results
10% → 3%
Forecast error, best products
“I was told our products are un-forecastable. This is a new way of looking at things - it’s too good to be true.”
— VP Supply Chain

Where to start?

POC on your own data.
See results in days.