Freemium Token Model Analysis: Are Global Platforms Affordable?

This project analyzes the effectiveness and fairness of token-based freemium models (like Upwork, Lovable, etc.) that give limited free credits to users.
We evaluate whether these models encourage daily active usage or unintentionally cause drop-offs — especially for users from low-income regions.

---

Problem Statement

Platforms like Upwork and Lovable use a freemium + credit/token model:

• New users get limited free tokens (e.g., connections or interactions)
• Once exhausted, users must pay to continue using core features

Many users, especially from developing countries, sign up but drop off after hitting these limits.

Key question:
Is this model affordable and sustainable across different income regions?

---

Project Objectives

1. Analyze token pricing & usage rules of global platforms
2. Collect real-world income & cost-of-living data per country
3. Simulate user behavior (free use, upgrade, churn)
4. Build an Affordability Index: can users afford to stay?
5. Generate recommendations for global freemium pricing

---

Data Lifecycle

We follow a 7-step data science lifecycle:

1. Business Understanding – Define the problem and hypotheses
2. Data Collection – Real-world token pricing + income stats
3. Data Preparation – Clean, transform, and integrate datasets
4. EDA – Visualize user flow, pricing pressure, and country-wise costs
5. Modeling – Simulate churn and engagement patterns
6. Evaluation – Check fairness and friction across regions
7. Deployment – Generate PDF report + publish case study

---

Data Sources

Dataset	Description	Type	Source/Status
pricing.csv	Token models and pricing from real apps	Real	Manually compiled
income.csv	Country-wise income and cost of living	Real	Numbeo, World Bank
user_behavior_simulated.csv	Simulated free vs. paid user behavior	Simulated	Built in Python

---

🧰 Tools We Will Use

• Python: pandas, seaborn, matplotlib, sklearn
• Jupyter Notebook: for EDA, modeling, and simulation
• Excel / Google Sheets: for manual entry and pricing matrices
• GitHub: to version control and publish our project
• Notion: to document hypotheses and ideas
• LinkedIn: to publish the final case study for professional exposure

---

Data Ethics & Assumptions

• This project uses simulated and publicly available data.
• We do not represent or misuse any internal company data.
• Token models are inferred based on user experience and publicly visible pricing.
• Simulated behaviors are designed for analytical purposes only.

---

Why This Project Matters

Freemium token-based pricing is common in SaaS, freelancer platforms, and AI tools.
But many users in developing countries experience friction — they run out of credits quickly and can't afford subscriptions.

We aim to answer:
Are these platforms globally inclusive or accidentally exclusive?

By combining real-world income data, platform pricing, and simulated usage patterns, this project provides insights that go beyond dashboards.
We recommend business strategies that balance monetization with accessibility.

---

Status

• Phase 1: Project Setup – Completed
• Phase 2: Data Collection – Ongoing
• Phase 3: EDA & Affordability Analysis
• Phase 4: User Behavior Simulation & Churn Modeling
• Phase 5: Reporting, Recommendations, and LinkedIn Case Study