The situation
The manufacturer needed to adapt production capacity quickly to trend shifts, which required accurate forecasts of aggregated sales volumes.
Forecasting was built mainly in Excel, and indicators were chosen by correlation, so many had no causal effect on sales.
The challenge
- Minimise the forecast errorExcel-based forecasts with irrelevant indicators gave poor accuracy.
- Understand the marketIdentify new market drivers, and evaluate new policies and their impact on the commercial vehicle market.
- More marketsExtend coverage to more markets, for both demand and production.
What Indicio did
- Found a seasonal pattern in sales, low in January, rising through spring to a first peak in May, and removed it before modelling.
- Used a Lasso model to test the combinations of indicators and suggest the optimal group, instead of correlations in Excel.
- Applied 50+ econometric models and weighted them into one forecast.
The result
- 40.4%
- lower forecast error 12 months ahead: MAPE 4.68% against 7.86%
- 1–2 months
- earlier detection of market trends
- 351.8K
- vehicles per year forecast, against 352.8K actual (internal forecast: 339.6K)
The 12-month ahead forecast error fell from 7.86% to 4.68%.
New and stronger leading indicators detect market trend changes one to two months earlier.
A structured process across all markets, with significantly stronger management support.
How it was done
Seasonal patterns
The data was cleaned and the seasonal pattern identified and removed before modelling, then added back to the forecast.
Benchmark forecasts
Univariate models built on historical sales alone set the benchmark that the more advanced models had to beat.
Specific leading indicators
A VAR model with a Lasso penalty tested the possible combinations of indicators and kept the group that best predicts future sales.
Multivariate forecasts
A library of multivariate models forecast the main variable together with the indicators identified.
Weighted by accuracy
Some models are better at short, mid or long horizons. Indicio combines them into one forecast, giving more weight to the best model at each step.