In-Store Visitor Behavior Analysis to Improve Visitor Engagement & Sales

Case Study:

The Client:

Dxilogy worked with Digital Mortar on this case. The target site was a large Department Store with 25+ departments/sections. Navigation and engagement data were collected for 25,000 unique visitors in the Physical store.

Learn about machine learning and store optimization from Diginal Mortar CEO, Gary Angel. Click here to watch.

The DXi Solution:

Using beacon apps or in-store sensors, it was possible to capture a large volume of data regarding in-store visitor movement and engagement across different departments, and across different product aisles within the same department. Visitor movement to the checkout counter (Cash Wrap) or exit from the store without approaching the checkout counter, was captured in the same manner.

The data was then treated in the same manner as data collected from website visitors would be treated. Using the DXi algorithm plus other Machine Learning algorithms, a probability-of-conversion (make a purchase), as well as the DXi score (experience), was generated for each in-store visitor.

The DXi Platform identified the top factors which contributed to the engagement and conversion (Cash Wrap visit) of each in-store visitor.

Top factors included visitor behavior attributes (e.g. time spent in store) as well as attributes such as “Specific Products”, “Specific Sections (e.g. Backpacks, Casuals)” and “Locations (e.g. fitting rooms).

DXi provided dimensionality for deeper segmentation of the in-store visitors by grouping and classifying based on their behavior pattern within the store. Analysis of the visitor navigation and engagement data identified in-store behavior patterns which would lead to higher visitor engagement and hence higher sales. The DXi Platform is able to handle constantly varying visitor engagement patterns, which are difficult to address by traditional business rules.

Quote from the Client, Gary Angel, Founder / CEO, Digital Mortar:

“DXi brings the power of advanced machine learning to bear on your in-store shopper behavior data. DXi can take your Digital Mortar data and produce best path analysis, shopper clustering and so much more. It’s the most advanced way to optimize your store layout and merchandising”

The Results:

Actionable Insights

In-store behavior pattern indicting high-probability conversion of a visitor

  • Visit Time > 796 seconds
  • No of Associates > 2
  • Specific area = Hoodies

Product Categories / Products indicating higher in-store engagement

Most important:

  • Backpacks
  • Casual Shoes
  • Men’s Jackets
  • Running Shoes
  • Women’s Pants

Least Important:

  • Women’s Tights (most surprising as this was a prominent display in central area of the store)

Store Sections leading to higher engagement and sales:

  • Higher Value for Fitting Rooms
  • Lower Foyer value

Benefits and Next Steps

Probability of conversions would be increased by:

  • Engaging a visitor to more than 796 seconds, mostly usefully by ensuring number of times an associate meets each visitor is more than 2
  • Focus on Hoodies section and make it more prominent, and display in Foyer and Cash Wrap sections to enhance conversions/sales
  • Prioritization / ranking of departments which enhance the engagement of visitors.
  • Re-assign merchandize to location occupied by Women’s Tights

With Fitting Rooms and Foyer sections playing a key role with the engagement of the visitors the store should focus on improving the visitor experience on these 2 locations.

Cash Wrap section is also critical.