I will build a customer churn prediction model with dashboard
About this Gig
Struggling to know which customers are about to leave before they actually do?
I build churn prediction models and customer segmentation reports that show you exactly who's at risk and why, so you can act before you lose them.
What I've done: Built a churn prediction system for an e-commerce retailer on 541,909 real transactions flagged £379K in at-risk revenue using RFM segmentation and a Random Forest model with 94% recall (catching 94 of every 100 customers who were about to churn).
What you get, depending on your package:
- RFM customer segmentation see who your Champions, Loyal, At Risk, and Lost customers are
- A trained churn prediction model not just a notebook, but a working tool
- An interactive Streamlit app for real-time churn scoring
- A Power BI dashboard for prioritizing retention budget (Premium)
- Clear explanations of why customers are churning, not just a risk score
Before you order: message me a sample of your data (or describe what you have customer ID, purchase dates, order values are the key columns needed) so I can confirm scope and give you an accurate timeline.
Programming language:
Python
•
R
•
MATLAB
•
SQL
•
Colab
Frameworks:
Scikit-learn
•
Panda
Tools:
Jupyter Notebook
•
Excel
•
Colab
•
RStudio
FAQ
What format should my data be in?
CSV or Excel works best. At minimum I need customer ID, transaction/order date, and order value — more columns (product category, region, etc.) can improve accuracy.
Do you need my data to be clean already?
No — data cleaning and preprocessing is part of every package.
Can you work with a small dataset?
Yes, though results are more reliable with at least a few hundred customers and multiple transactions per customer (needed for RFM analysis).
What if I don't know what "churn" means for my business yet?
I'll help define it during scoping — e.g., "no purchase in 90 days" — based on your typical purchase cycle.
Is the source code included?
Yes, in Standard and Premium packages.

