User Growth Strategies: From Information Management to a Data‑Driven Flywheel

Introduction The author, a data professional, introduces user‑growth methods by referencing Borges' "The Garden of Forking Paths" and the concept of random walks, emphasizing the need to navigate uncertainty and find stable paths for growth.

1. Information Management: From Chaos to Order Before the internet, computers served scientific calculations. Early MIS (Management Information Systems) evolved from material management to enterprise resource planning, gradually turning chaotic data into organized information.

2. Information Distribution: From Inefficiency to Efficiency Initially, information was listed in yellow pages. With the internet explosion, search engines emerged, ranking pages based on quality and traffic. Later, recommendation systems appeared, using user‑item interactions to present the most relevant content.

3. Information Production: From Complexity to Simplicity In recommendation and search pipelines, after retrieval, information must be recognized, understood, and integrated. With the rise of large models (AIGC), many production steps can be automated, simplifying the workflow.

4. Product as Entropy Reduction Product evolution reduces entropy: information management moves from disorder to order, distribution from low to high efficiency, and production from complexity to simplicity. This frees users from chaotic experiences and creates a sense of security.

5. Data‑Driven Flywheel The author describes a "data flywheel" where data consumption links business and data wheels, forming a growth engine. Examples include AB‑testing for user retention, ByteDance’s data‑driven culture (OKR, dashboards), and the construction of agile tools (data application layer, data asset layer, engines) to support rapid decision‑making.

Case study: a short‑video platform increased playback by 300% through two interventions—adding a short‑video card in the homepage recommendation flow (100%+ increase) and improving the feed‑down module with new‑user guidance (140%+ increase).

6. User‑Growth Stages Stage 1.0 focuses on acquisition and activation (DAU) via ads or content marketing. Stage 2.0 emphasizes retention, treating it as the north‑star metric. Stage 3.0 treats growth as a continuous loop, forming a business flywheel that optimizes key loops and sustains long‑term value.

7. Q&A Q1: How to evaluate a metric in AB testing and avoid Simpson’s paradox? A: Perform cohort analysis, ensure attribution is clear, and recognize that global results may mask subgroup effects. Q2: What is DataFinder? A: An interactive analysis tool with built‑in models (e.g., retention) that lets users click‑through analyses without writing SQL. Q3: How should data‑science teams collaborate with business? A: Work in agile, cross‑functional squads, share metrics, and give data scientists the authority to set experiment standards.

Conclusion User growth must be built on a product that delivers real value; only then can sustainable commercial returns be achieved.