Recategorize products using network analysis & clustering models

Grouping products with similar relationships together

Support in-store and in-app digital assortment layouts

Visualize category hierarchy that represents customer decision making

Visualize category hierarchy that represents customer decision-making

Improve other analytics that use product categories as an input (segmentation, CLV key drivers, etc.)

Key Concept

Impact

Quantify the relationship or “lift” between two products using apriori rules mining

Calculate lift, support, and confidence, and combine with chi-squared statistical tests

Identify bundling opportunities for products with high lift

Use as a feature for halo/cannibalization effect for promotion effectiveness

Predict interest score for every customer x product combination

Separate models can be built for new products or customers with no data

Find top potential customers for product-specific promotions

Combine with churn prediction for personalized retention campaign

In this complex world,

we need to collaborate to deliver outstanding work.