Business to Business do not use demographics

Final Project - Data Analytics Ironhack

Clustering a B2B Client Database

On B2B besides the direct commercial relationship, valueable information can be obtained using VAT numbers.

• Current status of the company.
• Declared main activity of the company.
• Location.
• Company P&L.

This information could be helpful to: Develop a look-alike process to obtain new clients Find out accounts under/over-perfoming compared to their peers Find out buying patterns in your B2B clients database

Technology Stack

SQL, python, pandas, selenium, matplotlib, UMAP, HDBSCAN

Core

Extracting client’s buying history from the Company database. Classified every gsale on a product-category. Through webscrapping at einforma.com enrich the dataset with valuable client information:

• Declared main activity
• Location Province
• Current operativity status

Reduce the dataset with UMAP and proceed to clustering with HDBSCAN.

After the clustering, a csv file is generated with all the data that can be analyze using a Business Inteligent Software like Power BI or Tableau.

Folder Structure

└── project
    ├── .gitignore
    ├── requeriments.txt
    ├── README.md
    ├── notebooks
    │   ├── notebook1.ipynb
    │   └── notebook2.ipynb
    ├── final_data
    └── images