Situation

Our client is a global powerhouse in the food and beverage industry where product placement has a direct impact on sales. Sales reps are tasked to manually count items on shelves and display racks, monitor inventory, and ensure compliance. This process was time-consuming and prone to errors. Additionally, reporting was delayed, since reps had to return to the office to input their findings.

Our AI tool increases productivity, improves display compliance, and reduces out-of-stock situations.

Sales reps needed a solution to monitor compliance more efficiently, track out-of-stock items, and optimize product assortments. Meanwhile, company leadership sought robust tools for competitive intelligence and demand forecasting to drive brand growth.

Action

We implemented an AI tool using imaging and video that sales reps can use to compare in-store displays with digital planograms for compliance. Sales reps access an app on their mobile devices, to take before-and-after photos of the shelf or display. These images are audited using computerized visual processing technologies and degree of compliance is evaluated by combining spatial and product information and comparing the location and product types against the planogram. Compliance and stock levels can now be viewed in real time.

When using video, objects of relevance, such as a can or rack, noting its shelf position are detected and identified. Based on different attributes of the objects and specific business rules derived from the data, KPIs and other data are calculated, generating a report instantly.

In addition, the AI tool determines missing products, recommends products for the shelf display, provides demand forecasts, identifies key competitors; and evaluates product pricing and discounts.

Our solution saves time for sales reps, improves display compliance, and reduces out-of-stock situations.

Results

  • Boosted revenue and sales due to product pricing and optimization opportunities.
  • Improved product mix, accuracy and shelf compliance using AI image recognition.
  • Increased productivity and accuracy for sales reps that take inventory in multiple stores.
  • 500K

    Images processed in a year and 200K events coded for modelling
  • 90%

    Accuracy in model
  • 5K +

    Stores recorded weekly