Marketing mix modelling

Marketing mix modelling is the econometric technique that estimates from historical data how much revenue each marketing and environmental factor contributed: media, price, promotion, distribution, seasonality and competition. Unlike digital attribution it works without personal data and also sees offline channels and longer-term effects, which makes it the standard for top-level budget allocation. The flip side: a model is only as good as its data and assumptions, needs years of variation, and produces wide confidence intervals that tend to disappear in management presentations. Sensible use pairs model results with validation experiments and treats the output as direction, not accounting truth.

Covered in