Bundling Analysis

Key Highlights

  • Potential to increase revenue by ~ $4.7M ($100K - $250K per location) by bundling together “recommended products” to “existing kits” in 40 locations (**based on 2019 kit volume)
  • Understand customer purchasing behavior and changes over a period of time
  • Alignment with customer purchasing patterns, offering a simplified purchasing experience

Challenges

The client was using a traditional process of bucketing the items/products into bundles which is not gaining traction with their customers as they are going for more ad-hoc contracts in certain locations. The objective is to develop an intelligent bundling strategy leveraging customer purchasing patterns and restructure their current bundles/create new ones to gain customer attention and increase sales volume

Approach

  • Analyzed the client’s existing bundling plans/strategy to understand the gaps and identify the areas of improvement
  • Designed and developed an unsupervised learning model market association rule mining that looks for hidden patterns in transactional data and recommends the list of products that can be bundled together
  • The association rules//results can be leveraged by business using tableau dashboard
  • Adopted Kanban methodology to track and achieve day-to-day deliverables

Tech Stack

Python, MSSQL, Jupyter notebook, Machine learning –Market Basket Analysis algorithm (apyori), TFS, JIRA

Related Case Studies