Course overview
Business plans, budgets, supply commitments and investment decisions depend on forecasts that can be explained, tested and revised as new evidence arrives. Spreadsheet-based projections often obscure assumptions, fail to handle multiple drivers consistently and become difficult to audit. Python provides a practical environment for joining operational, financial and market data; producing repeatable forecasts; measuring forecast error; and testing management scenarios under stated assumptions.
This course teaches participants to build business forecasting and scenario-analysis workflows using Python. Participants prepare time-series data, explore trend and seasonality, establish benchmark forecasts, and apply methods including moving averages, exponential smoothing, ARIMA, regression and Prophet. They assess forecast accuracy with MAE, RMSE, MAPE and bias, select models against business requirements, and convert driver assumptions into base, upside and downside scenarios. The course also addresses hierarchical forecasts, intermittent demand, forecast governance, model monitoring and communication for non-technical decision-makers.
Instruction combines concise technical briefings with worked cases based on sales planning, cash forecasting, demand planning and workforce capacity. Participants write and review Python notebooks, interpret model output, test scenario assumptions and present forecast recommendations. They leave with a reusable forecasting notebook structure, validation checklist, scenario model and reporting template that can be adapted to their own datasets after the course.
The course suits analysts and managers who already work with business data and need to move from static projections towards documented, repeatable forecasting processes. It is particularly relevant to finance, commercial, operations, supply-chain and planning teams responsible for forecasts that influence decisions.
