Select Page

Retail Store Traffic Analysis

DPS Retail Store Traffic Analysis is a Computer Vision based in-store analytics solution for offline retailers using Cameras to capture metrics like Footfall, Demographics, Conversion Rate, Busy Areas in store, Dwell Time, etc.

Retail Store Traffic Analysis utilizes already installed cameras to gather data like, Number of People Entered, Number of People Passed by the store, Time spent in an Area of store, Most Visited Areas of Store, and etc. The application uses the video feed from cameras and applies Detection and Tracking of People moving In or Out of the store, or an Area of the store.

The application gathers this data to produce reports using Business Intelligence Models, on the provided Web-Interface where the Store Owner can review these reports and understand the factors that can benefit the store as well as those that are affecting the store. The data gathered consists of following:

1. Store (How busy the store is inside and outside)

  • Entries (Total People that enter the store)
  • Passer-by(s) (Total people that pass-by the store)
  • Capture Rate (Percentage of people that enter the store compared to those that pass-by)
  • Average Occupants (Average number of people in the store)
  • Average Dwell Time (Average amount of time people spend in the store)

2. Area (How busy each area of the store is)

  • Total Time (Total time all people spend in an area)
  • Average Time (Average time a person spends in the area)
  • Entries (Total people that enter the area)
  • Maximum Occupants (Maximum people in the area at the same time)
  • Occupancy (Percentage of time the area is occupied by at least one person)
  • Average Occupants (Average number of people in the same area at the same time, when it is occupied by at least one person)
  • Entry Rate (The amount of times a person entering the store enters different zones inside the store)
3. Queue (How busy Queues are)

  • Average Time (How long on average a person spends in the queue)
  • Maximum Occupants (Maximum people in a queue at the same time)
  • Average Occupants (Average people in the queue at the same time, when it is occupied by at least one person)
  • Frequency (Percentage of time there is a queue with at least one person)
4. Sales (Point of sale metrics)

  • Conversion Rate (Percentage of people that enter the store and then make a purchase)
  • Transactions (Total number of purchase transactions)
  • Volume (Total value of all sales)
  • ATV (Average Transaction Value)

Technology Stack:

  • Web-Interface
    • JavaServer Faces
    • PrimeFaces
    • Hibernate
  • ML-Interface
    • Python 3.7
    • OpenCV
    • DLIB

Benefits:

  • Measure business critical metrics to keep your stores performing optimally.
  • Keep track of the busiest times in-store, from footfall at the entrance to queue times at the till.
  • See how your stores could be performing better — measure peel-off rates outside your store to improve window displays arrangements.
  • Know which customer demographics to target with localized digital campaigns, in-store media and digital displays, to drive traffic optimally to the stores
  • See which area in store takes longest to serve customers and adjust staffing schedule accordingly.
  • Reporting tools using Computer Vision can show you which shopper demographics are most likely to convert, or which times of day have increased service times that may be impacting sales performance.