Multi-Touch Attribution Model and Marketing Budget Optimization

Business Objective

Marketing attribution is essential for understanding the contribution of various marketing channels to overall success. This project focuses on multi-channel attribution modeling to quantify the value of advertising campaigns and enhance advertising ROI. The primary goal is to empower stakeholders to make data-driven decisions by isolating the impact of each touchpoint on conversions.

---

Aim

This project aims to develop and compare various attribution models to identify the marketing channels that drive higher customer conversions and optimize budget allocation for maximum efficiency.

---

Data Description

The dataset used in this project is in CSV format, comprising 586,737 rows and 6 columns. The columns include:

• Cookie: Anonymized customer ID.
• Time: Date and time of the customer visit.
• Interaction: Type of interaction (categorical variable).
• Conversion: Binary indicator (0 for not converted, 1 for converted).
• Conversion Value: Monetary value of the conversion event.
• Channel: The marketing channel responsible for directing the customer to the site (target variable).

---

Tech Stack

• Programming Language: Python
• Libraries:
  • NumPy for numerical computations.
  • Pandas for data manipulation.
  • Matplotlib and Seaborn for data visualization.
  • Itertools for iteration utilities.
  • Gekko for mathematical optimization.
  • Pandas-Profiling for generating automated EDA reports.

---

Approach

1. Import Dependencies:

   • Load all required libraries and modules for the analysis.

2. Data Import:

   • Read the dataset into a Pandas DataFrame.

3. Exploratory Data Analysis (EDA):

   • Generate a comprehensive EDA report using Pandas-Profiling.
   • Perform data cleaning and visualization to understand key trends and patterns.

4. Build Single Touch Attribution Models:

   • Last Touch Attribution Model: Assign credit to the last channel before conversion.
   • First Touch Attribution Model: Assign credit to the first channel in the customer journey.
   • Last Non-Direct Touch Attribution Model: Assign credit to the last channel before conversion, excluding direct visits.

5. Build Multi-Touch Attribution Models:

   • Linear Attribution Model: Distribute credit evenly across all touchpoints.
   • Position-Based (U-Shaped) Attribution Model: Assign 40% credit to the first and last touchpoints, and 20% to the intermediate points.
   • Position Decay Attribution Model: Assign decreasing credit to touchpoints as they get further from the conversion event.

6. Build Probabilistic Attribution Models:

   • Markov Attribution Model: Use transition probabilities to evaluate channel contributions.
   • Shapley Value Model: Apply cooperative game theory to determine the marginal contribution of each channel.

7. Results:

   • Consolidate results from all models into tables for comparison.
   • Generate and save visualizations to illustrate model outputs.

8. Budget Optimization Engine:

   • Use optimization algorithms (e.g., GEKKO) to allocate marketing budgets efficiently across channels based on model results.

---

Project Structure

.
├── input/                                # Contains input dataset (e.g., attribution_data.csv).
├── src/                                  # Source code folder with modularized scripts.
│   ├── engine.py                         # Main script to execute the pipeline.
│   ├── ML_Pipeline/                      # Folder containing modular Python functions.
│       ├── data_processing.py            # Data preprocessing utilities.
│       ├── attribution_models.py         # Attribution modeling functions.
│       ├── optimization.py               # Budget optimization functions.
├── output/                               # Stores output visualizations and reports.
├── lib/                                  # Includes the original Jupyter notebook and reference materials.
├── ppt/                                  # Presentation slides for project explanation.
├── requirements.txt                      # File containing project dependencies.
└── README.md                             # Project documentation.

---

Getting Started

1. Clone the Repository:

git clone <repository_url>
cd <repository_folder>

2. Install Dependencies:

Use the following command to install all required libraries:

pip install -r requirements.txt

3. Run the Project:

Navigate to the src folder and execute the engine.py script:

python src/engine.py

4. Explore the Outputs:

• Check the output/ folder for model graphs and reports.
• Review detailed EDA insights and model results.

---

Results

• Attribution Insights:
  • Quantified the contribution of marketing channels to conversions.
• Budget Optimization:
  • Reallocated budget for improved ROI.
• Visualizations:
  • Clear and actionable graphs to support decision-making.

---

Contributing

Contributions are welcome! To contribute:

1. Fork the repository.
2. Create a feature branch:

git checkout -b feature-name

3. Commit your changes:

git commit -m "Add feature"

4. Push your branch:

git push origin feature-name

5. Open a pull request.

---

License

This project is licensed under the MIT License. See the LICENSE file for more details.

---

Contact

For any questions or suggestions, please reach out to:

• Name: Abhinav Navneet
• Email: [email protected]
• GitHub: AjNavneet

---

Acknowledgments

Special thanks to:

• GEKKO for providing optimization capabilities.
• Pandas-Profiling for automated EDA insights.
• The Python open-source community for excellent tools and libraries.