Recommendation Engine

Recommendation Engine: Increase Upselling and Cross-Selling Opportunities for Retail and F&B

The consumer industry is changing, with customers shopping online and using digital purchase channels such as QR ordering more than ever before. Retailers and F&B outlets need to leverage on these digital channels to target the right products to the right customers, helping customers discover products in a relevant and timely way.

With IMDA’s Retail Recommendation Engine (RRE), retailers and restauranteurs can increase sales by offering tailored recommendations to customers. The recommendation engine uses machine learning algorithms, which can be trained on vendor-specific data, to find the right products for customers based on preferences, browsing behaviour, and trends.

• Product Similarity feature helps recommend products that are similar to what they are currently viewing. If a customer is browsing the men's shirts section, RRE would recommend a similar shirt of a different colour or design.
• Product Association recommends products based on what other shoppers have purchased with the item they are currently viewing. If a customer is looking at a pair of jeans in the store, RRE would suggest other items that people have purchased alongside those jeans, like a belt or shoes, to help them complete their look.
• Personalisation generates recommendations based on each customer's individual purchase history and preferences. For example, if a customer has purchased items from one in the past, they will likely be interested in future purchases from another associated category. This feature allows us to recommend products when we know it would be relevant for that customer.
• Trending identifies trending searches and purchases across all customers so that retailers can stay ahead of trends.

The RRE is available at no cost for all solution providers that develop systems for local retailers and F&B outlets. The model is built in a manner that makes it easy to integrate as is it fully containerised, and all it takes to integrate, train, and customise RRE is a few lines of code.

Interested solution providers are welcome to reach us at AI@imda.gov.sg.