Method
Markov chains / HMM
Transition-matrix models of states and behavior.
9 cases from 9 companies apply Markov chains / HMM. 1 was checked in an online A/B test; 3 report results in production.
Typical tasks: Evaluate models, Explain a metric change, Analyze user behavior, Measure impact without an A/B test and 7 more. Companies: Netflix, Positive Technologies, Avito, Expedia Group, Duolingo, trivago and 3 more. Often combined with: LLM-as-a-judge, Synthetic data generation (LLM), LLM simulation / role-play, Clustering, Synthetic control and 19 more.
Analyzing Public Pedagogical Benchmarks for AI Tutors and Building Synthetic Students for Testing![]() | ![]() | edtech | 2026-05 |
Beyond Funnels: Advanced UX Analytics![]() | ![]() | general | 2026-05 |
| Part 2: A Survey of Analytics Engineering Work at Netflix | media-streaming | 2025-01 | |
Hacker Hunting: Analyzing Process Launch Graphs![]() | ![]() | general | 2024-11 |
| How We Moved from a Markov Chain to a SASRec Model | classifieds | 2024-05 | |
| Marketing Attribution | travel | 2023-04 | |
| Meaningful metrics: How data sharpened the focus of product teams | edtech | 2023-02 | |
| Marketing Attribution: Evaluating The Path to Purchase in the Product Ecosystem | travel | 2022-12 | |
| 4 Essential Steps for Building a Simulator | food-delivery | 2022-08 |


