Retail Decision Modeling

IBM

The Challenge

Are buying decisions based on how long a person sees an object? How does time in the store and placement of items affect purchasing decisions? We worked with IBM on a project that has been the holy grail for retailers for years: Does proximity to an item encourage a purchase?

Our Insight

The folks at Big Blue had developed a 360° fisheye camera and software that can track the movements of people through a store over time. We wanted to capture and model this data within a geospatial context.  

Our Approach

We entered the process by creating the data-fabric of a trial store: all the aisles and shelves and products. We then were able to process the data coming from IBM’s tracking camera and created a histogram comparing the time spent in front of products to sales of those items, using the data from IBM’s point of sale database.

Categories

  • Data
  • ESRI
  • Geodatabase
The indoor analytics of people moving through space and time has allowed IBM to test and analyze various methods to increase customer conversion rates, as well as improve store layouts and presentations.
The challenge was to determine how time in the store and placement of items affects purchasing decisions.
We worked with IBM on a project utilizing modeling and analysis in a retail environment, to determine whether proximity to an item encourages a purchase.

The Results

The indoor analytics of people moving through space and time has allowed IBM to test and analyze various methods to increase customer conversion rates, as well as improve store layouts and presentations.

Explore our related projects