Beverages — Premium Spirits
Reduction in forecast error
Reduction in expedited (airfreight) shipping cost
About
A premium spirits maker needed to balance ageing-barrel commitments made years in advance with fast-changing market demand today. Flagship whiskies vanished from 820 outlets while slow-moving RTDs filled warehouses, tying up Rs. 52 crore in cash and forcing expensive airfreight. We built a Benders-decomposition model that schedules production and distribution while tracking barrel evaporation.
Industry
Beverages — Premium Spirits
Company size
1,000 – 5,000 employees
Founded
1962
The Company
A premium spirits maker planning years ahead and days ahead at once
The company produces aged whiskies alongside faster-moving ready-to-drink products, a portfolio that forces two very different planning horizons to coexist: barrel-ageing decisions made years in advance, and market demand that shifts month to month.
Getting the balance wrong in either direction is expensive — an under-supplied flagship whisky means lost sales in a category with real brand cachet, while an oversupplied RTD line means warehouse space and working capital tied up in product that isn't moving.
The challenge
Flagship stock-outs and RTD overstock at the same time
Flagship whiskies were vanishing from 820 retail outlets even as slow-moving RTD products filled warehouse space, tying up Rs. 52 crore in working capital. The mismatch was forcing the business into expensive airfreight to plug gaps that a longer-horizon plan should have anticipated.
The core difficulty was that ageing-barrel commitments couldn't be planned the same way as fast-moving RTD demand, and the business had no single system reconciling the two — production and distribution decisions were effectively being made on separate timelines that never properly talked to each other.
The Solution
A Benders-decomposition model that tracks barrel evaporation alongside demand
We built a Benders-decomposition optimisation model that schedules production and distribution jointly while explicitly tracking barrel evaporation and ageing curves — treating long-horizon ageing commitments and short-horizon demand as one connected planning problem rather than two separate ones. Monte-Carlo dashboards let planners rehearse sourcing and ageing strategies before committing to them.
The model surfaces the trade-off directly: what a given ageing-stock allocation implies for near-term availability, and what a near-term demand shift implies for future ageing-stock planning — letting planners see both horizons on the same screen instead of reconciling them manually.
The Results
30% better forecasts and airfreight costs down 32%
Forecast error fell 30% as production and distribution planning finally worked from the same model rather than separate assumptions. Expedited airfreight shipping costs — previously the fallback for every planning miss — dropped 32%, and flagship-whisky availability was restored across the 820 outlets that had been experiencing stock-outs.
Shelf availability rose broadly across the portfolio as the model's ability to rehearse sourcing and ageing scenarios in advance meant fewer surprises reaching the distribution stage in the first place.
KEEP READING
Explore more studies on AI, analytics, and enterprise intelligence.
