Series
Three arcs, one goal: turn a business question into a defensible number.
Most posts here belong to one of three ongoing arcs. Each arc has its own reading order, so you can start from the beginning or jump to what you need.
Pick an arc from the Series menu at the top of the page. If you’re new here, start with Causal Inference — it’s the foundation the other two build on.
🟢 Causal Inference
Identification strategies, potential outcomes, DAGs, and the methods that let you defend a number in front of a skeptical stakeholder.
Covers: A/B testing, DiD, RDD, IV, synthetic control, matching, G-computation, DML, causal forests, uplift modelling.
🟠 Pricing & Demand
Price elasticity, two-sided market pricing, dynamic pricing, MMM, and the causal experiments that make pricing decisions defensible.
Covers: elasticity estimation, personalized pricing, MMM, geo experiments, revenue optimization, marketplace dynamics.
🔵 Forecasting
Probabilistic forecasting, model evaluation, and where forecasting stops being causal inference. From ETS to foundation models, with honesty about what each can and cannot do.
Covers: ETS, ARIMA, hierarchical reconciliation, probabilistic forecasting, conformal prediction, foundation models.
How the arcs connect
The three arcs are separate but they answer complementary questions.
- Causal inference asks what would have happened otherwise?
- Pricing asks what should we do?, and reuses causal methods to measure the effect of price and marketing decisions.
- Forecasting asks what is likely to happen?, and feeds both pricing decisions and experimentation planning.
New posts publish weekly. If you want them in your inbox, subscribe via the index.xml or drop me a message on LinkedIn.