Today brand teams have many different channels and tactics at their disposal for building brand equity and boosting sales. However, with the increasing complexity of the promotional landscape it becomes critical to make data-driven decisions about how much to invest in each channel.

Promotional Mix Modeling (PMM) is a technique to quantify the impact of advertising and promotional activity on the sales of a brand while controlling for other factors such as product price, distribution, category growth, and competition. PMM informs marketing strategy and tactics by helping brand teams understand the effectiveness of each marketing channel such as advertising, in-store, direct-to-customer (DTC), digital and social media.

Client is a CPG manufacturer with a portfolio of leading beverage brands. To promote their Target Brand, Client spends $10+ MM each year allocated across diverse promotional channels such as In-store, TV Ad, Print Ad, Paid Search, and Social Media.

Client needs to measure the Return on Investment (ROI) on each channel to understand which promotions are more effective in driving sales.

Business data for brand sales, pricing, promotional investment etc. was extracted on a weekly basis for 3 years. A longitudinal dataset was stitched together comprising of sales, promotions, and control variables

The Regression module within the MLAutomata platform was used to model the relationship between brand sales and promotions such as Television and Print Advertising, Digital Paid Search and Social Media. Variables representing brand price, in-store promotions, seasonality etc. were included in the model as controls. A Multivariate Regression model was used to estimate the impacts. Finally, ROI (return on investment) for each channel by applying brand margin, promotional investment and other financial parameters.