Skip to main content
Case StudyAI← All case studies

27% less waste with per-flavor forecasting.

An ice cream manufacturer replaced a one-size-fits-all forecasting approach with model selection for each flavor, reducing waste and saving planning time.

27%Reduction in production waste
7hManual planning time saved monthly
Explore the results
The result

Less production waste. Less manual planning.

27%

Reduction in production waste

Reduction in production waste reported after introducing per-flavor forecasting.

7h

Manual planning time saved monthly

Manual production planning time recovered each month.

About these results

These outcomes are reported in the published Software Sushi case study. The published account does not include the baseline dataset or measurement methodology.

  • Reduction in production waste: Published case study · Results
  • Manual planning time saved monthly: Published case study · Results
The challenge

One forecast could not serve every flavor.

Business
Ice cream manufacturer
Data
Historical production and sales
Work delivered
Per-flavor forecasting and backtesting

The client was overproducing and underproducing across their flavor lineup, leading to significant waste and lost revenue. A single forecasting model couldn't capture the distinct seasonality and demand curves of each flavor.

Before

A single forecasting approach missed the distinct demand patterns of each flavor, leading to overproduction and underproduction.

After our work

Each flavor used its best-performing model, selected by backtesting against historical data.

Our work

Select a forecasting model for each flavor.

The approach

Select models at the flavor level. Historical production and sales data provided the basis for backtesting five candidate architectures per flavor, rather than applying one model to different demand patterns.

The solution path

Production planning tailored to each flavor’s seasonality and demand.

Conceptual workflow · based on the published case study
  1. Model by flavor

    Built a dedicated forecasting model for each individual flavor

  2. Compare candidates

    Evaluated 5 different model architectures per flavor

  3. Backtest

    Backtested all models against historical production and sales data

  4. Select

    Selected the best-performing model per flavor — no one-size-fits-all

Let's talk

Does demand vary across your product range?

Let’s discuss your forecasting process, historical data and production-planning challenges.

Book a Free Consultation