Loading...

Use Case: 'People Like You'

Affinity and association rule learning based online recommendation systems, such as market basket analysis with Apriori-algorithm, are well understood and widely used in the e-commerce environments. However, when these methods are applied in furniture retail several challenges emerge. While it is relatively easy to collect significant amounts of data and make accurate predictions on a single item for the frequently purchased consumer items, like books and apparel, the furniture retail recommendation systems face exactly same challenge as that "traditional" furniture salesman: lack of information about the customer's home environment and personal styling. Whilst some things, like the most popular fabric choices and combinations can be recommended with over a 80% probability, without an understanding of the customer's home (room size, items, colour schemes) it is close to impossible to go beyond this, let alone "fill the room" with your products.

VividPlatform changes the way recommendations can be made. When the customer creates a model of his/her home (possibly with photos) the size, colour schemes and other items in the room are known and the recommendations can be generated based on the real customer environment (e.g. association rules between other shoppers' behaviour, environment, and interior designer "lookbook"). As the information is permanently stored in the retailer's systems and integrated to the CRM, the same recommendation engine can also be used for generating targeted and relevant marketing to the consumers, e.g. recommend a new sofa table that fits the customer living room and matches the colour schemes and styles, enabling the retailer's brand to remain relevant to the consumer between all their purchases.

Leave a Comment