Use Product Recommendations - But Be Smart About It.

A good recommender system must cater to and cover a large number of shopping journeys. It exists to alleviate a merchandiser from the heavy load of constantly tweaking a manual system, especially in high SKU count brands.

Think about the following use cases that can be fully automated: 

• Inspiring New Customers to begin a shopping journey, for example with a Geo-Location based ‘Trending in Your Area’ Recommendation. 

• Reinforcing existing interests from hesitant window shoppers, for example by using a Netflix inspired ‘Continue Shopping’ Recommendation 

• Supporting Product Discovery for existing customers who have already left signals to the website and expect new exciting products, for example by using a browsing history related recommendation

• Driving higher baskets and cross-sells, for example by using bought together or visually similar AI recommendations 

• Upselling customers at the point of highest intent directly in Checkout with Post-Purchase Upsell Recommendations.

Bonus points if the recommendations are personalized based on signals such as size, brand, category or price point affinity. 

Expected uplift is 3-15% divided between higher CR, AOV and Retention Rate