This project was completed as part of the Quantium Data Analytics Virtual Internship.
The objective of this project is to evaluate the effectiveness of a new in-store layout by comparing the performance of trial stores with carefully selected control stores using historical sales data.
Using statistical analysis and data visualization, the project measures whether the new trial layout generated a significant uplift in store performance and provides business recommendations based on the results.
The Category Manager wanted to determine whether the new trial store layout should be rolled out across all stores.
To answer this business question, each trial store was compared against a carefully selected control store with similar historical performance before the trial period.
- Selected suitable control stores using Pearson Correlation.
- Applied Magnitude Distance to identify the closest matching stores.
- Compared trial and control store performance.
- Performed statistical significance testing.
- Created business-ready visualizations.
- Generated data-driven business recommendations.
- Select suitable control stores using historical sales data.
- Compare trial stores against matched control stores.
- Measure sales uplift during the trial period.
- Perform statistical analysis.
- Visualize business insights.
- Provide recommendations for future rollout.
| Trial Store | Selected Control Store |
|---|---|
| 77 | 233 |
| 86 | 155 |
| 88 | 237 |
- Load and prepare the dataset.
- Calculate monthly performance metrics.
- Select suitable control stores using:
- Pearson Correlation
- Magnitude Distance
- Compare trial stores with selected control stores.
- Perform statistical significance testing.
- Generate visualizations.
- Summarize findings.
- Provide business recommendations.
| Category | Technology |
|---|---|
| Programming | Python |
| Data Analysis | Pandas, NumPy |
| Visualization | Matplotlib |
| Statistics | SciPy |
| Development | Jupyter Notebook |
| Version Control | Git & GitHub |
Quantium_Task2_Experimentation_and_Uplift_Testing
│
├── data/
├── notebooks/
│ └── Quantium_Task2_Experimentation_and_Uplift_Testing.ipynb
├── output/
│ ├── charts/
│ │ ├── store77_trial_assessment_sales.png
│ │ ├── store86_trial_assessment_sales.png
│ │ └── store88_trial_assessment_sales.png
│ └── trial_store_summary.csv
├── README.md
├── requirements.txt
├── LICENSE
└── .gitignore
The analysis identified the following control stores:
| Trial Store | Control Store |
|---|---|
| 77 | 233 |
| 86 | 155 |
| 88 | 237 |
- Store 77 showed a positive sales uplift during the trial period.
- Store 86 did not show a statistically significant uplift.
- Store 88 showed a positive sales uplift during the trial period.
Overall, the trial layout produced positive results in two out of three trial stores.
Based on the statistical analysis:
- Store 77 demonstrated a positive uplift.
- Store 88 demonstrated a positive uplift.
- Store 86 did not show a statistically significant improvement.
The new store layout appears to be successful in the majority of the trial stores.
It is recommended to consider a broader rollout of the new layout while further investigating the factors affecting the performance of Store 86 before a full implementation.
git clone https://github.com/syed-ubedullah/Quantium_Task2_Experimentation_and_Uplift_Testing.gitcd Quantium_Task2_Experimentation_and_Uplift_Testingpip install -r requirements.txtnotebooks/Quantium_Task2_Experimentation_and_Uplift_Testing.ipynb
Execute all notebook cells from top to bottom to reproduce the analysis and visualizations.
https://www.linkedin.com/in/syed-ubedullah-basha
GitHub
https://github.com/syed-ubedullah
This project was completed as part of the Quantium Data Analytics Virtual Internship for learning and portfolio development purposes.
The project demonstrates experimentation, uplift testing, statistical analysis, business decision-making, and data visualization using Python.