Theseus: A Python Library for Cohort Analysis

Theseus serves as a specialized open-source Python library engineered to streamline cohort analysis, retention profiling, and product growth modeling within the digital gaming and software sectors. Its primary purpose is to automate the complex mathematical task of fitting raw retention data to established growth models, including logarithmic, exponential, and power functions. By identifying the most accurate statistical fit for historical user behavior, the library enables stakeholders to generate reliable long-term projections regarding user retention and platform sustainability.

The library provides a robust framework for calculating Daily Active User (DAU) trajectories based on incoming user cohorts. Beyond simple forecasting, it offers advanced analytical capabilities such as age-based user segmentation and the ability to reverse-engineer acquisition requirements. By inputting specific DAU targets, users can determine the precise volume of new user acquisition necessary to achieve their growth objectives. This functionality transforms raw retention metrics into actionable strategic insights, allowing product teams to align acquisition efforts with long-term retention realities.

Practical implementation of these tools facilitates the aggregation of cohorted DAU data, enabling a comprehensive view of how individual cohorts contribute to the total active user base over time. The library further supports operational workflows through integrated export features, allowing for the seamless transfer of projection data into Excel or JSON formats for broader organizational reporting. By standardizing the methodology for retention modeling, Theseus provides a scalable solution for data-driven decision-making across the product lifecycle, ensuring that growth projections remain grounded in empirical user behavior.