In this project, our task was to identify major customer segments on a transnational data set with 55,000 records which contains all the transactions occurring between 01/12/2010 and 09/12/2011 for a UK-based and registered non-store online retail.The company mainly sells unique all-occasion gifts. Many customers of the company are wholesalers.
Customer segmentation is the process of grouping customers together based on common characteristics. These customer groups are beneficial in marketing campaigns, in identifying potentially profitable customers, and in developing customer loyalty.Once you have these segments, you can build the right product, set the right distribution and positioning, and match the right sales motion to each customer, while also refining your segments over time. Done well, it’s a model that gives anyone at your company an immediate understanding of our customers.
-
Demographic segmentation
-
Geographic segmentation
-
Behavioral segmentation
-
Psychographic segmentation
-
Customer journey segmentation
Once you have these segments, you can build the right product, set the right distribution and positioning, and match the right sales motion to each customer, while also refining your segments over time. Done well, it’s a model that gives anyone at your company an immediate understanding of your customers.
- Online Retail Customer Segmentation.ipynb - This file includes Features description, exploratory data Analysis, feature engineering, feature scaling and implemented algorithms.
- Customer Segmentation PPT - This is a power point presentation file of a project. It includes various visualaized plots of EDA using Seaborn and Matplotlib. The result chart of various implemented algorithms.
- Project Summary - A brief summary of aim and methodology of the project.
- InvoiceNo: Invoice number. Nominal, a 6-digit integral number uniquely assigned to each transaction. If this code starts with letter 'c', it indicates a cancellation.
- StockCode: Product (item) code. Nominal, a 5-digit integral number uniquely assigned to each distinct product.
- Description: Product (item) name. Nominal.
- Quantity: The quantities of each product (item) per transaction. Numeric.
- InvoiceDate: Invice Date and time. Numeric, the day and time when each transaction was generated.
- UnitPrice: Unit price. Numeric, Product price per unit in sterling.
- CustomerID: Customer number. Nominal, a 5-digit integral number uniquely assigned to each customer.
- Country: Country name. Nominal, the name of the country where each customer resides.
- RFM Analysis: https://www.analyticsvidhya.com/blog/2021/07/customer-segmentation-using-rfm-analysis/
- Kmeans: https://towardsdatascience.com/k-means-clustering-algorithm-applications-evaluation-methods-and-drawbacks-aa03e644b48a
- Agglomerative Clustering: https://www.javatpoint.com/hierarchical-clustering-in-machine-learning
We have engineered features to obtain new features such as RFM, RFMGroup, and RFMScore for getting more details of customers' purchasing behaviour.We have employed use of unsupervised learning algorithms KMeans and Hierarchical Agglomerative clustering to segment the customers into 2 major groups.We have achieved Silhouette score of 40% and generated optimal number of clusters for best and worst costomers.
Consumer segmentation enables us to uncover valuable audience insights that can help to dictate the content, targeting and media buying strategy of our digital campaigns. It allows us to understand which audiences exist, who to target and why, and the most effective way to reach them.We have grouped our wholesale customer base into two major segments one comprising best customers with high recency frequency and monetary scores while other on the lowest side.This will help us in formulating marketing strategies and build the right product to boost sales.
- Fabien Daniel | https://www.kaggle.com/code/fabiendaniel/customer-segmentation/notebook
- Sanjay Yadav | Data Scientist
https://www.shopify.in/encyclopedia/customer-segmentation
https://scikit-learn.org/stable/modules/generated/sklearn.cluster.KMeans.html