The analysis aims to enhance dataset quality through thorough preprocessing, including outlier management, handling missing values, and resolving duplicates. It focuses on creating customer-centric features like Recency, Frequency, and Monetary (RFM) metrics, along with other behavioral insights such as average purchase intervals and peak shopping times, to uncover meaningful transaction patterns. This approach facilitates comprehensive customer segmentation and activity-level categorization for actionable insights in CRM analytics.
- You can access the complete project python file here - [Python]
- You can access the complete project in pdf format here - Report
The dataset encompasses transactions from 01/12/2010 to 09/12/2011 for a non-store online retail business based and registered in the UK. Specializing in distinctive all-occasion gifts, the company's clientele includes a significant number of wholesale customers.
Product Portfolio Specializing in distinctive all-occasion gifts, the company's clientele includes a significant number of wholesale customers.
| Feature | Description |
|---|---|
| InvoiceNo | Invoice number that consists 6 digits. If this code starts with letter 'c', it indicates a cancellation. |
| StockCode | Product code that consists 5 digits. |
| Description | Product name. |
| Quantity | The quantities of each product per transaction. |
| InvoiceDate | This represents the day and time when each transaction was generated. |
| UnitPrice | Product price per unit. |
| CustomerID | Customer number that consists 5 digits. Each customer has a unique customer ID. |
| Country | Name of the country where each customer resides. |
Scope of Business Expansion
- The main market of the firm is in United Kingdom but there are transactions with customers from other countries also. There is a large scope of expansion, for the first step, if Germany and France are having even 5% of the customers from UK, the additional revenue will be about 1 Million more per year which is 11% of the current total revenue.
Tailored Promotional plans
- 'Hibernating' segment is large in both number and revenue. They are less frequent customers with low order value but offering relevant products and special discounts will create brand value again and bring them up to promising customers. This can boost the revenue by approximately 1 Million per year with us 11% of the current revenue. -There is a good percentage of revenue in 'At Risk' segment. This section of customers are important since they contribute to about 10.4% of the revenue in the given time period. It will be good to bring those customers back by sending personalized emails to reconnect with special offers and sharing useful resources with them. -There is a section of customers in 'Needing Attention' segment. Their contribution to the revenue is 5.6% in the time period but they became inactive towards the end of the period. Launch limited-time offers and recommendations based on previous orders to reactivate them.
Welcome offers and Promotions for New Customers
- There are promising number of 'New customers' in the given time period. But the average order value of them is less. Introduce the new customers to promotions and offers. The new customers have a potential to increase the revenue by 20% in the future i.e, by 1.78 Million per year.