In-Store Product Recommendation System Using Location-Aware Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Physical retail stores face challenges in providing personalized product recommendations due to limited customer information and the inability to associate recommendations with the customer's location within the store, leading to inefficient sales experiences.
Innovation Solution
A computer-implemented system that uses facial recognition, mobile phone communication, and indoor positioning systems to identify customers, retrieve user-specific information, and generate product recommendations based on their purchase history, location, and expressed interests, with rules stored in memory to guide merchants to recommended products using augmented reality and visual indicators.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If facial recognition and mobile phone communication are used to identify customers and retrieve user-specific information, then personalized product recommendations can be generated, but device complexity and implementation cost increase
Solution Approach 1:
The system segments customer identification into multiple independent modules: facial recognition module, mobile phone communication module, and user profile retrieval module. Each module handles a specific function, allowing the system to achieve comprehensive personalization while maintaining manageable complexity through modular architecture.
Solution Approach 2:
The system employs universal identification mechanisms that can work through multiple channels (facial recognition, mobile phone communication). These multi-functional identification methods can serve various purposes including customer recognition, authentication, and preference tracking, reducing the need for separate systems for each function.
2Adaptability or versatility
If indoor positioning systems are implemented to track customer location, then location-specific product recommendations can be provided, but device complexity and implementation cost increase
Solution Approach 1:
The system uses an intermediary positioning infrastructure consisting of beacons deployed throughout the store and an indoor positioning system that translates physical location into digital coordinates. This intermediary layer enables location-specific recommendations without requiring complex direct tracking between customers and products, as the positioning system acts as a mediator that maps store space to recommendation contexts.
3Productivity
If augmented reality and visual indicators are used to guide merchants to recommended products, then sales efficiency improves, but device complexity increases
Solution Approach 1:
The system creates a virtual overlay (copy of the physical store view) through augmented reality that displays recommendation information. Instead of requiring complex physical modifications to the store or products, the system generates a digital copy of the store environment with superimposed guidance indicators, enabling efficient product location while maintaining simple physical infrastructure.
Data Source
AI summary
When a person visits a physical retail store, the merchant often does not have enough information about the person to make meaningful product recommendations. Also, a physical retail store typically has products physically distributed throughout the store. It may be desirable to have some sort of relationship between where the customer is and the location of the product being recommended. In some embodiments, when a person visits the store, a computer determines an identity of the person and generates a product recommendation based on user-specific information for that person. In some embodiments, generating the product recommendation includes detecting that a field of view of a camera of the device has changed, and in response determining a plurality of products within or proximate to a current field of view of the camera. At least one of the plurality of products is then identified as the recommended product.