Tags: Retail, Image Processing, Product Identification, Brand Identification

Trax/Nielsen Shelf Analytics

Having the right product on the shelf is a critical pillar in many retail strategies and tactics. That begins with knowing what is on the shelf.

Historically shelf analytics was done by detail “men” that would periodically come into a store, representing some product group, and assess the state of the shelf for product placement (products pay for optimal locations) and to check for products not available (the efficacy of store restocking). With traditional retailers struggling against the e-Commerce giants, improvements have been made in shelf analytics systems (e.g., the image processing and analysis discussed here), and much more is possible (and necessary) going forward.

The Trax/Nielsen Shelf Analytics Suite is a collaboration between Trax, a very well capitalized ($400M) Singapore-based retail solutions vendor, and ACNielsen, the US headquartered market research firm. Trax provides optimization and analysis solutions for retailers; Nielsen contributes industry benchmarks and comparisons, and integration with other market analytic offerings. The current version of the suite processes photos taken of retail shelves replacing granular, manual input from the detail person.

Industry: Retail

Business Need: Understand the efficacy of shelf optimization strategies (e.g. End Cap purchases of prime spots) as well as the ability of a store to abide by their commitments and restock shelves effectively.

Key Enabling Technologies: Image Processing, Product Identification, Brand Identification

Industry Perspective

Ideally a product company would like to do shelf analytics based on near real-time data updates (and integrate with visitor tracking and behavior analysis).

Edge Need: The real-time analysis is computationally intensive. Edge services offer an attractive alternative to installing and operating in-store servers.

Ease of Incorporation: These advanced services would blend well with advanced retail strategies. Technically, unlike visitor tracking, dynamic analysis requires additional and newly located cameras, or the use of robotic cameras (an alternative that is being explored).

Be a part of the Edge experience.

Share Your Edge Story