Client Situation
•A major super market chain sought new store locations with the highest potential for expanding sales.
•While F&B sales continued to grow, client had not kept up with the broader market.
•Client sought answers to key questions:
Where do we under/over perform relative to the competition?
Where should we open up new retail outlets?
What is the demand profile of shoppers at existing retail locations?
Results
•Delivered 1.2% sales increase in current locations.
•Identified geographic areas for expansion of new store openings.
•Developed quantitative insights to guide shelving, assortment and merchandising decisions.
SFC VALUE-ADD
We analyzed store sales data across different retailers and times. It was not a surprise to discover that while the food and beverage category continued to grow, our client along with other traditional grocery retailers were declining. But we found the decline was not uniform across all locations and markets.
We relied on Artificial Intelligence predictive modeling techniques to project a proprietary shopper demand segmentation onto our customer's loyalty card database. This mapping allowed us to establish the demand characteristics for individual existing locations and evaluate them against the broader market.
Depending on the competitive position in the marketplace and the demand potential for any given store, we were able to partition the market into clusters with different allocations of investments. Our strategy resulted in 1.2% sales increase in current locations.
Classifying locations outside of our existing footprint required that we construct different types of predictive algorithms since we lacked individual household-level data beyond our current locations. We used credit bureau data, magazine subscriptions, mail order activity, hobbies and interests, among many other fields, supplied by third-party data suppliers, to build and apply predictive models for individual households. Once we typed all 190 Million households in the US, we aggregated them up to a proprietary store trading areas to map the demand profiles to locations outside of our customer's existing retail footprint.
Given that we had been to measure demand for every household, after the aggregation to block groups, we were in a position to test our own new store trade areas. It was possible to execute custom geo-fencing by rolling up block groups to new locations. Furthermore, we could evaluate any new prospect location against our existing store clusters and determine how to allocate resources for the execution of precise expansion plans.
