Objective: To perform a variety of customer analytics, including customer segmentation, calculating customer lifetime value and predicting next purchase date for an ecommerce business, and to provide value-add recommendations based on the analytics.
Tools Used: Python, Pandas, scikit-learn, Matplotlib, Seaborn
Skills Demonstrated: Data Cleaning, Feature Extraction, Exploratory Data Analysis, K-Means Clustering (Unsupervised Learning), Linear Regression (Supervised Learning), Customer Segmentation, Data Visualization
Recommendations:
Based on the analysis performed, the ecommerce business needs to consider taking the following actions:
- Run promotions in Q1 each year to increase the number of active customers and revenue, to avoid the drop off experienced each year.
- Focus marketing efforts on attracting new customers and increasing the customer base.
- Conduct an investigation to determine the reasons for the peaks and valleys in monthly average revenue per order.
- Investigate why average revenue per order decreases in December each year, when it would be expected that there is high customer demand and averages should be increasing.
- Investigate why average revenue per order decreases in May or June each year. It may be that a different product mix is required at that time of year.
- Consider stocking more products with higher selling prices, likely increasing margins.
- Strongly consider growing its business in other European markets. Marketing efforts should be increased in these other markets, to increase the customer base and to grow revenue.
- Target specific marketing efforts towards the Low and Mid segments to generate more repeat buying. It is significantly more cost effective to market towards existing customers than it is to new customers.
- Conduct a survey to determine why many customers do not repeat buy.
- Reduce reliance on a small group of customers, by increasing repeat buying of existing customers and attracting new customers to the business.