Incorporating holistic Marketing Mix Models to optimize DME spends and improve sales for a leading FMCG company
01. Business Challenge
02. Approach
Analytics Methodologies
BAYESIAN REGRESSION MODEL
In a Bayesian framework, linear regression is stated in a probabilistic manner. That is, the Frequentist linear regression model is reformulated to use probability distributions. The syntax for linear regression in a Bayesian framework looks like this:
In essence, the response data points y are sampled from a multivariate normal distribution that has a mean equal to the product of the β coefficients and the predictors, X, and variance of σ2. Here, I refers to the identity matrix, which is necessary because the distribution is multivariate.
03. Impact
Key insights and recommendations from the analysis helped client improve their allocation strategy and plan the yearly budget
DME has significant impact on Volume gain and Preference gain (Brand Equity), with every 10.0% increase in DME results 10.9% increase in Volume. Hence the client was able to optimally allocate promotional spend across channels as part of the annual budget cycle.
Key Highlights
- Determining optimal allocation of DME spend
- Use of Bayesian Regression Model
- Non-Linear Transformation to capture volatility