This project uses R and D3 to show how a treemap can be used to explore and analyze product classification (also known as product taxonomy) for ecommerce sites.
[Treemaps] (https://en.wikipedia.org/wiki/Treemapping) are good at showing how properties are distributed through a hierarchy. Here are some classic examples:
- showing how disk space is used on your filesystem
- showing how market capitalization and price changes are distributed across sectors, subsectors, and individual stocks in a portfolio or stock exchange.
Large ecommerce sites tend to classify their products using a hierarchy of categories (taxonomy), to help with site navigation and faceted search. For example on Etsy there are 38M products with over 2000 different product classification in a hiearchy that is 6 levels deep. For catalogs and product taxonomies, a treemap can help answer questions like:
- How are my products distributed?
- Are my categories granular enough?
- Too many products in a category = can be hard to find what you're looking for.
- Too few products in a category = your store feels sparse.
- Which categories are the most popular / get the most traffic?
- Which categories have the highest conversion rates?
- How has a metric (e.g. product count/margin/sales/conversion) for a category changed over time?
Here is a simple example using Etsy's taxonomy. Area represents the number of products in a category, and color is mapped to the top-level category. This set of screenshots starts with the distribution of products across top-level categories, and using the treemap to drill down into the Hats and Caps category.
