Optimizing AB InBev’s Customer Segmentation with ML

Business Situation

AB InBev wanted to increase sales by improving segmentation of potential prospects and determining the appropriate portfolio of products to present to a given customer. To do so, the company needed to develop an application that would identify new prospects for their sales representatives. Maven Wave was engaged to build an analytic framework and mapping tool for AB InBev derived from current craft beer data and data from the Google Maps APIs.

Solution

Maven Wave integrated Google Maps, Places, and AB InBev market data to define customer segments and prioritize sales efforts using a variety of geospatial and machine learning techniques.

The application required three key outputs: Maven Wave would comprise the data into a Portfolio Fit Index; the information would be rendered dynamically on a map; and the data needed to be exportable.

After Maven Wave gained a better understanding of AB InBev’s current business challenges, the scope of the project included a Machine Learning (ML) model based on a variety of factors including zip code, the number of reviews on Yelp, price level and other key data. Maven Wave also developed an algorithm that could scan a defined outer region and grid search areas.

Results

The company’s application enabled AB InBev to identify what ‘tier’ a prospect might fall into, with various tiers representing possible revenue spend. The solution provided AB InBev sales representatives not only with additional customer prospects to target, but a prioritized list of top-tiered prospects. The customer has since expanded the application to include additional cities.