Learning

Causal Inference for the Brave and True

Matheus Facure's comprehensive Python-based coverage of synthetic control, difference-in-differences, and other causal methods central to marketing science.

Matteo Courthoud's DiD Tutorial

Industry perspective with full Python code. Covers classic DiD with potential outcomes, parallel trends testing, multiple time periods, Card-Krueger replication, and business applications.

Mixtape Sessions GitHub Repository

Free workshop materials from sessions taught at Facebook, eBay, LSE, and Oxford. Covers advanced DiD, staggered timing, PT violations, with coding labs and interactive apps.

Asjad Naqvi's DiD Repository

The definitive meta-resource for modern DiD. Covers TWFE failures, Goodman-Bacon decomposition, all major estimators (Callaway-Sant'Anna, Sun-Abraham, etc.) with code in Stata, R, Python, and Julia. Updated quarterly.

Pedro Sant'Anna's DiD Resources

14 lecture slide decks from the co-creator of Callaway-Sant'Anna. Covers classical DiD, parallel trends, ML for DiD, event studies, TWFE problems, and treatments turning on-and-off.

Jonathan Roth's DiD Resources

Course slides and coding exercises focusing on pre-trends testing limitations and HonestDiD sensitivity analysis. Created the HonestDiD and pretrends R packages. Includes practitioner checklists.

Google's CausalImpact Blog Post

Production-grade tool from Google's advertising team. Bayesian structural time-series approach with automatic variable selection and uncertainty quantification. Widely used for marketing impact analysis.

Stitch Fix: Market Matching with CausalImpact

Industry application combining dynamic time warping with CausalImpact for marketing intervention analysis. Shows how synthetic control concepts are adapted for real business problems at scale.

Causal Inference for the Brave and True: Time Series

By economist at Nubank. Chapters 13-15, 24-25 address panel data/time series causal analysis. DiD, synthetic controls, RDD with time dimension. Bridges econometrics and ML with executable notebooks.

Matteo Courthoud's Synthetic Control Tutorial

SCM for industry practitioners with references to Google, Uber, Facebook use cases. Python implementation with sklearn and cvxpy. Explains SCM as 'transpose of regression' with placebo inference.

Scott Cunningham: Causal Inference Substack

Substack and podcast 'The Mixtape with Scott' featuring interviews with leading causal inference researchers. Bridges academic methods and practical application.

Causal Inference: The Mixtape

Scott Cunningham's academic-quality but accessible methodology covering causal methods essential for marketing measurement.

Double/Debiased Machine Learning Guide

From the original DML authors. Explains Neyman orthogonality, cross-fitting, DML with text/complex data. Focuses on practical implementation rather than theory.

Awesome Causal Inference (Matteo Courthoud)

Comprehensive GitHub repository curating causal inference resources. Papers, packages, tutorials, and datasets organized by topic.

First Course in Causal Inference (Python)

Python implementation of Peng Ding's textbook 'A First Course in Causal Inference'. Educational resource with code examples.

Causal Econometrics Course

Graduate-level credibility revolution methods. Comprehensive coverage of modern causal inference techniques for econometricians.

Google Research: CausalImpact Paper

Foundation paper for CausalImpact package: inferring causal impact using Bayesian structural time-series models for interrupted time series.

Causal Inference: A Statistical Learning Approach

Stefan Wager's free PDF textbook covering causal inference from a machine learning perspective with theoretical foundations and practical applications.

Causal Wizard Reading List

Organized learning path for causal inference. Quality-assessed progression from basics to advanced methods.

Mark White's Practical Causal Forest Tutorial

Explains why optimize directly on causal effects, not outcomes. Complete workflow from data prep to interpretation using GRF package. Written for applied researchers transitioning to causal ML.

Stanford ML & Causal Inference Short Course

Video lectures from Susan Athey, Jann Spiess, and Stefan Wager covering ML vs. econometrics, ATEs with propensity scores, CATE estimation with causal forests, and loss functions for causal inference.

Uber Engineering: Uplift Modeling for Multiple Treatments

Extending X-Learner and R-Learner to multiple treatments with cost optimization. Production system design for uplift models at scale with cost-aware treatment allocation.

AEA: Machine Learning and Econometrics (Athey/Imbens)

9 hours from two of the most influential computational economists. ML vs. causal inference, heterogeneous treatment effects, LASSO/random forests, causal forests, policy learning. Athey pioneered ML in economics; Imbens won 2021 Nobel.

Applied Causal Inference Book (Chernozhukov et al.)

Comprehensive online textbook covering DML, causal forests, and modern causal ML methods with Python/R code. Essential reference for practitioners.

LOST Statistics: Causal Forest Tutorial

Practical guide to implementing causal forests for heterogeneous treatment effect estimation with code examples in R and Python.

DoorDash: Causal Modeling to Get Value from Flat Experiment Results

Extracting value from neutral experiments via CATE estimation using S-learner and T-learner. When overall effects are null, heterogeneous effects may still exist.

Stanford STATS 361: Causal Inference Lecture Notes (Wager)

Stefan Wager's graduate course notes on causal inference with machine learning: heterogeneous treatment effects, conformal inference, and forest methods.

Matteo Courthoud's IV Tutorial

IV in experimental settings with realistic tech examples (newsletter subscription as instrument). Covers LATE/Compliers interpretation, exclusion restriction, weak instruments diagnostics. Complete Python code.

Matteo Courthoud's RDD Tutorial

RDD fundamentals, bandwidth selection methods, and replication of Lee, Moretti, Butler (2004). Practical implementation with Python code using statsmodels.

Andrew Heiss's RDD Course Examples

Complete sharp vs. fuzzy RDD comparison with downloadable datasets. Shows rdrobust() usage, 2SLS with iv_robust(), and compliance visualization. Reproducible R code with tidyverse.

Tilburg Science Hub RDD Tutorials

Based on Cattaneo, Idrobo & Titiunik. Covers ITT vs. LATE, monotonicity, bandwidth selection for fuzzy designs, and multi-dimensional RDD. Includes Colombian education subsidy replication.

Uber Engineering: Causal Inference at Uber

Real industry application showing how PhD-level methods translate to business problems. Covers propensity score matching at scale, RDD for dynamic pricing, and mediation modeling.

Teconomics: Machine Learning Meets Instrumental Variables

How to reframe past A/B tests as instruments for behaviors you cannot randomize. Covers IV for behavioral effects, Deep IV, and ML for instrument selection. Actionable for data scientists.

Causal Inference for the Brave and True: Time Series

By economist at Nubank. Chapters 13-15, 24-25 address panel data/time series causal analysis. DiD, synthetic controls, RDD with time dimension. Bridges econometrics and ML with executable notebooks.

Andrew Heiss: Synthetic Data for Program Evaluation

Comprehensive guide for creating synthetic data for DiD, RDD, and IV evaluation designs. Includes R code examples.

Spotify: Encouragement Designs and Instrumental Variables for A/B Testing

IV estimation for experiments with noncompliance. How to handle experiments where users don't follow their assigned treatment, using complier populations.

Microsoft Research: Adversarial ML and Instrumental Variables

Innovative approach combining adversarial machine learning with instrumental variables for flexible causal modeling in complex settings.

Andrew Heiss's DAG and Backdoor Tutorials

Hands-on tutorials on building DAGs with ggdag, backdoor criterion, confounders/colliders, d-separation, and propensity scores. Uses real variable names with complete R code.

Brady Neal's Introduction to Causal Inference

14-week video course covering potential outcomes, DAGs, do-calculus, and causal discovery. Features guest lectures from Susan Athey, Alberto Abadie, and Yoshua Bengio. Bridges ML and econometric traditions.

Nick Huntington-Klein: Animated Causal Graphs

Innovative visual demonstrations of how different causal methods work. Animated DAGs showing confounding, selection bias, and identification strategies.

Lyft: Causal Forecasting at Lyft

Two-part series on DAG-based structural modeling and causal forecasting for marketplace decisions at Lyft.

R-causal Book: DAG Construction Chapter

DAG construction with ggdag. Practical guide to building directed acyclic graphs for causal inference in R.

PyWhy: Causal Discovery Example

PC, GES, LiNGAM algorithms for discovering causal structure from data. When you need to discover the causal graph rather than assume it.

DoorDash: Using Back-Door Adjustment for Pre-Post Analysis

Causal graphs and covariate adjustment for pre-post analysis. How to properly control for confounders when randomization isn't possible.

Adam Kelleher: Causal Data Science Medium Series

Former BuzzFeed data scientist's accessible series on graphical causal inference. 'If Correlation Doesn't Imply Causation, Then What Does?' and more.

Matteo Courthoud's Meta-Learners Tutorial

S-learner, T-learner, X-learner with detailed math, causal trees/forests, AIPW estimators. Uses Uber's CausalML package for demos. Complete Jupyter notebooks on GitHub.

KDD 2021 Tutorial: Causal Inference with EconML and CausalML

Industry workshop with 4 case studies from Uber, TripAdvisor, Microsoft. Ready-to-run Google Colab notebooks covering uplift modeling, customer segmentation, and long-term ROI estimation.

Netflix: Computational Causal Inference

Technical deep-dive into Netflix's causal inference infrastructure, software tools, and scalable computation approaches for causal analysis.

Adam Kelleher: Causality Python Package

Python implementation of causal inference algorithms including do-sampler, causal graph inference, and conditional independence testing.

First Course in Causal Inference (Python)

Python implementation of Peng Ding's textbook 'A First Course in Causal Inference'. Educational resource with code examples.

PyWhy: Causal Discovery Example

PC, GES, LiNGAM algorithms for discovering causal structure from data. When you need to discover the causal graph rather than assume it.

Microsoft Research: End-to-End Causal Inference at Scale Demo

Demo video showcasing Microsoft's EconML ecosystem for production causal inference, from data prep to deployment.

LinkedIn: Building Inclusive Products Through A/B Testing

Novel approach to measuring inequality impact of experiments. How to ensure product changes don't disproportionately harm certain user groups.

Practical Introduction to Switchback Experiments

Tutorial on designing and analyzing switchback experiments for marketplace experimentation.

EGAP Learning Days: Field Experiments

Theory and practice of field experiments with international development focus. Randomization, power analysis, and ethical considerations.

021 Newsletter: Marketing & Data Teams Bridge

Barbara Galiza bridging marketers and data teams (7,000+ subscribers). When to use click attribution vs MMM, how to structure incrementality testing, marketing data infrastructure.

Data Analysis Journal (Olga Berezovsky)

Weekly newsletter bridging academic statistics and product analytics practice. Experimentation guides, A/B test checklists, and workflow best practices.

Netflix Tech Blog: What is an A/B Test?

Multi-part series covering metric selection, sequential testing at scale, quasi-experimentation when SUTVA is violated, and interleaving for recommendation testing. Published at KDD.

Ronny Kohavi: Seven Rules of Thumb for Web Site Experimenters

Foundational paper establishing core practices for online experimentation. Rules still followed at major tech companies today.

Marketing Analytics (UVA Darden/Coursera)

Rajkumar Venkatesan's course covering brand measurement, CLV, experiment design, and marketing resource allocation. Strong focus on causal inference for marketing.

150 Successful ML Models at Booking.com (KDD 2019)

Reveals that model performance ≠ business performance. Demonstrates why RCTs are critical for validating ML models in production with framework for hypothesis-driven iteration.

Netflix: A Survey of Causal Inference Applications

Comprehensive overview of how Netflix applies causal inference across experimentation, personalization, and content decisions at scale.

Wayfair: Geo Experiments for Incrementality

Convex optimization for treatment assignment when simple randomization won't work. Covers synthetic control matching and practical constraints like maximum geo share limits.

Haus Blog: Synthetic Control & Geo-Experimentation

PhD causal inference experts publishing rigorous content on geo-experiment fundamentals, synthetic control methodology, and why matched market tests are insufficient.

DoorDash: 4 Principles to Boost Experimentation by 1000%

Scaling from ~10 to 100+ experiments/month using switchback designs in logistics.

Spotify: Risk-Aware Product Decisions in A/B Tests

Framework for combining success, guardrail, deterioration, and quality metrics. How to make decisions when multiple metrics move in different directions.

Remerge Findings: Incrementality Testing Approaches

Technical breakdowns of incrementality testing methods from a DSP perspective. Covers intent-to-treat, PSA, ghost ads, and ghost bids with clear pros and cons.

Netflix: Page Simulator for Better Offline Metrics

Netflix Tech Blog on using simulation to test homepage recommendations before running A/B tests.

Ronny Kohavi: Trustworthy Online Controlled Experiments

The definitive book on A/B testing methodology by the architect of experimentation at Microsoft, Amazon, and Airbnb. 27,000+ citations.

Netflix: Return-Aware Experimentation

KDD 2025 Best Paper on optimal experiment design with limited resources. Framework for designing experiments that maximize learning given resource constraints.

GrowthBook's Experimentation Fundamentals

Complete single-page reference covering hypothesis formation, statistical significance, Type I/II errors, MDE, power analysis, A/A tests, novelty effects, and experiment interactions. Notes that industry success rates are only ~33%.

Uber: Analyzing Experiment Outcomes Beyond Average Treatment Effects

Quantile treatment effects for understanding distributional differences in marketplaces. Goes beyond ATE to measure how treatments affect different parts of the outcome distribution.

Netflix: Sequential A/B Testing Keeps the World Streaming

Anytime-valid inference at production scale. Real case study: detecting play-delay issues that would have prevented 60% of devices from streaming. Covers time-uniform confidence bands.

Evan Miller's A/B Testing Tools

Interactive calculators for sample size, chi-squared, sequential sampling, and t-tests. The companion article 'How Not To Run an A/B Test' is the canonical reference on why repeated significance testing inflates false positives.

Matteo Courthoud's Experimentation Series

Connects experimentation to econometric foundations. Covers CUPED (linking to DiD), group sequential testing, Bayesian A/B testing, and clustered standard errors. Every post includes complete Python code.

Evan Miller: Formulas for Bayesian A/B Testing

Mathematical foundations for Bayesian approaches to A/B testing. Derivations of exact formulas for posterior probabilities and expected loss.

Evan Miller: How Not To Run an A/B Test

The 250,000+ view article that shaped industry thinking on peeking problems. Essential reading on why continuously monitoring A/B tests leads to false positives.

Ron Berman: p-Hacking in A/B Testing

Wharton professor whose paper 'p-Hacking and False Discovery in A/B Testing' is critical reading. Demonstrates how common practices inflate false positives in marketing experiments.

LinkedIn: A/B Testing Variant Assignment at Scale

Hash-based variant assignment for trillions of daily invocations. Technical deep-dive on deterministic, consistent randomization at LinkedIn scale.

Spotify Confidence

Spotify's experimentation platform for feature flagging and A/B testing. SDK for controlled rollouts with built-in statistical analysis.

LinkedIn: Our Evolution Towards T-REX

Scaling to 41,000 simultaneous A/B tests on 700M+ members. How LinkedIn built infrastructure to support massive-scale experimentation.

Lukas Vermeer: Building Experimentation Infrastructure

Booking.com's Director of Experimentation on building culture and infrastructure for 1000+ concurrent experiments.

Spotify: New Experimentation Platform (Part 1)

Journey from ABBA to EP; Metrics Catalog for self-service analysis. Evolution of Spotify's experimentation infrastructure and lessons learned.

ExP Platform: Microsoft Experimentation Resources

Comprehensive experimentation platform guide from Microsoft's exp-platform team. Includes CUPED variance reduction, SRM detection, and metric design.

Airbnb: Experiments at Airbnb

Foundation article on Airbnb's experimentation platform: A/B testing infrastructure, metric design, and lessons from running experiments at scale.

Uber: Making Experiment Evaluation Engine 100x Faster

Engineering deep-dive on scaling experimentation infrastructure to 10M+ evaluations per second. Covers optimization techniques for high-throughput experiment analysis.

Netflix: It's All A/Bout Testing - The Experimentation Platform

Foundational overview of Netflix experimentation covering allocation, Ignite analysis tool, and monitoring. Architecture of one of the most sophisticated A/B testing platforms.

Matteo Courthoud's Experimentation Series

Connects experimentation to econometric foundations. Covers CUPED (linking to DiD), group sequential testing, Bayesian A/B testing, and clustered standard errors. Every post includes complete Python code.

Booking.com: Increasing Power with CUPED

Production-ready Hive SQL and Spark/R implementations for big-data scale. Handles missing pre-experiment data gracefully with real A/B test case study showing faster significance achievement.

Understanding CUPED by Matteo Courthoud

Mathematical derivation from first principles: optimal covariate formula θ = Cov(X,Y)/Var(X) and variance reduction Var(Ŷ_cuped) = Var(Ȳ)(1 - ρ²). Compares with DiD and Frisch-Waugh-Lovell theorem. Full Python code.

DoorDash: CUPAC for ML-Enhanced Variance Reduction

CUPAC (Control Using Predictions As Covariate) - ML-based CUPED extension for when standard CUPED fails. Achieved 25%+ reduction in switchback test duration.

ExP Platform: Microsoft Experimentation Resources

Comprehensive experimentation platform guide from Microsoft's exp-platform team. Includes CUPED variance reduction, SRM detection, and metric design.

Microsoft ExP: Deep Dive into Variance Reduction

From the team that invented CUPED. Comprehensive guide to variance reduction techniques for online experiments from Microsoft's Experimentation Platform.

Lyft: Challenges in Experimentation

Region-split tests with synthetic control and residualization for variance reduction. Advanced techniques for experimentation in ridesharing marketplaces.

Statsig's CUPED Deep Dive

Outstanding pedagogy using running/weights example. Demonstrates t-test and OLS regression equivalence, shows standard error reduction from 4.73 to 2.13, covers stratification approaches.

Eppo: CUPED++ for Extended Variance Reduction

CUPED++ extension using multiple pre-experiment metrics as covariates. Addresses 'new users have no pre-data' limitation. Quantifies impact: experiments can conclude 65% faster.

DoorDash: 4 Principles to Boost Experimentation by 1000%

Scaling from ~10 to 100+ experiments/month using switchback designs in logistics.

Practical Introduction to Switchback Experiments

Tutorial on designing and analyzing switchback experiments for marketplace experimentation.

Statsig: Switchback Experiments Overview

Best introductory resource with clear visual diagrams showing traditional A/B vs. switchback designs. Covers burn-in and burn-out periods to prevent cross-contamination.

DoorDash: Switchback Tests Under Network Effects

Why traditional A/B tests fail in three-sided marketplaces and how switchback testing with region-time randomization solves interference. Uses 30-minute time windows.

DoorDash: Statistical Analysis for Switchback Experiments

Deep methodology comparing OLS, Multi-Level Modeling, and Cluster Robust Standard Errors for switchback analysis. Addresses small independent units problem. Achieved 30% faster iterations.

Lyft: Experimentation in a Ridesharing Marketplace

Foundational article on SUTVA violations through potential outcomes framework. The bias-variance tradeoff table for randomization schemes (user to city level) is highly cited.

LinkedIn: Detecting Interference - An A/B Test of A/B Tests

Using cluster randomization to detect when user-level randomization causes interference. Methodology to test whether your experiments have network effects.

Meta: How Meta Tests Products with Strong Network Effects

Cluster experiments, power vs purity tradeoffs. How Facebook handles experimentation when treatment effects spill over between users.

Eppo: Bandit vs. Experiment Testing Decision Guide

The single best resource for when to use bandits vs. experiments. Covers perishable decisions, impact estimation challenges, why A/B tests win for complex multi-metric decisions.

Eugene Yan: Bandits for Recommender Systems

The definitive practitioner's guide synthesizing implementations from 12+ tech companies (Spotify, Netflix, Yahoo, DoorDash, Twitter, Alibaba, Amazon). Covers ε-greedy, UCB, Thompson Sampling.

Stitch Fix: Multi-Armed Bandits Experimentation Platform

Inside look at building bandit infrastructure. Covers Thompson Sampling convergence, deterministic allocation via hashing, and reward services architecture with feedback loop diagrams.

Evan Miller: Simple Sequential A/B Testing

Derives a simple sequential test using gambler's ruin: stop when T-C reaches 2√N. Elegant and implementable with basic arithmetic. Includes interactive calculator.

Matteo Courthoud: Group Sequential Testing

Pedagogical progression from peeking problem through Bonferroni, Pocock, O'Brien-Fleming to Lan-DeMets alpha-spending. Simulates 10,000 experiments showing Type I error rates. Full Python code.

Spotify: Choosing a Sequential Testing Framework

The definitive industry comparison of five frameworks: GST, mSPRT, GAVI, Corrected-Alpha, Bonferroni. Monte Carlo simulations comparing power. Maps methods to companies: GST (Spotify), mSPRT (Optimizely, Uber, Netflix).

Netflix: Sequential A/B Testing Keeps the World Streaming

Anytime-valid inference at production scale. Real case study: detecting play-delay issues that would have prevented 60% of devices from streaming. Covers time-uniform confidence bands.

Introduction to Agent-Based Modeling

Bill Rand's comprehensive free course on agent-based modeling from Santa Fe Institute's Complexity Explorer. Covers NetLogo, model design, and analysis.

ETH Zurich: Agent-Based Modeling of Economic Systems

GitHub repository with course materials for ETH's ABM course using Mesa. Includes exercises on market simulation and network effects.

Agent-Based Models with Python: An Introduction to Mesa

21-lesson course on Complexity Explorer covering agent-based modeling in Python using Mesa framework. Builds Sugarscape and other classic models.

Matteo Courthoud's BLP Demand Estimation

Exceptionally clear BLP from first principles. Share inversion, nested fixed-point step-by-step, instrument selection (BLP, Hausman, cost shifters), GMM estimation. Python implementation included.

Frank Pinter's Demand Estimation Notes

Builds intuition from multinomial logit → Berry (1994) → full BLP. MPEC vs. nested fixed-point, micro BLP with second-choice data. Written for PhD field exam prep with red bus-blue bus example.

Mesa Documentation Tutorials

Official Mesa ABM framework tutorials covering model building, data collection, and visualization step-by-step.

Econ-ARK DemARK Examples

Demonstration notebooks for HARK heterogeneous agent models. Includes buffer-stock, lifecycle, and Aiyagari model implementations.

QuantEcon Python Lectures

Comprehensive lecture series on computational economics covering dynamic programming, rational expectations, Markov chains, and heterogeneous agents.

RAND Research Archive

Free access to decades of defense policy research from the nation's oldest think tank, founded in 1948

Uber: Simulated Marketplace with ML

Agent-based discrete event simulation for testing dispatch algorithms safely. How Uber builds digital twins of their marketplace to test pricing and matching changes.

Lyft Engineering

Rideshare economics, forecasting, and marketplace efficiency. Technical deep-dives on pricing, dispatch, and causal inference.

Lyft: Simulating a Ridesharing Marketplace

Lyft engineering blog on counterfactual simulation framework for rideshare marketplace optimization.

Lyft: Quantifying Efficiency in Ridesharing

Efficiency isn't speed—it's an economic equilibrium. A masterclass in defining the objective function for marketplace optimization.

Lyft: Experimentation in a Ridesharing Marketplace

Foundational article on SUTVA violations through potential outcomes framework. The bias-variance tradeoff table for randomization schemes (user to city level) is highly cited.

Afi Labs: Ride-Share Dispatch Algorithms

Complete worked examples for ride-share dispatch with full code. Explains why greedy nearest-driver matching fails compared to optimal trip chaining.

Foundations of Transportation Network Analysis (edX)

MIT MicroMasters course on network modeling, traffic assignment, and transportation optimization. Part of the Transportation specialization on edX.

Class Central: Transportation Courses

Aggregated list of online transportation courses from universities worldwide. Filter by level, platform, and topic to find the right course.

MIT OCW: Transportation Systems Analysis

MIT's classic graduate course on demand modeling, networks, and intelligent transportation systems. Covers discrete choice theory, traffic flow, and transit planning.

Uber: Reinforcement Learning for Marketplace Balance

Largest RL deployment for matching in ridesharing—400+ cities globally. How Uber uses reinforcement learning to balance supply and demand in real-time.

Lyft: Challenges in Experimentation

Region-split tests with synthetic control and residualization for variance reduction. Advanced techniques for experimentation in ridesharing marketplaces.

MIT OCW: Energy Economics (14.44)

Paul Joskow's MIT course on theoretical and empirical perspectives in energy markets. Covers electricity, oil, gas, and environmental economics with full lecture notes.

Energy Institute at Haas Blog

UC Berkeley's Energy Institute blog featuring accessible research summaries on electricity markets, climate policy, and transportation. Written by leading energy economists.

MIT OCW: Engineering, Economics and Regulation of Electric Power

Interdisciplinary MIT course linking engineering, economic, legal, and environmental perspectives on electricity. Covers market design, reliability, and renewable integration.

MIT OCW: Energy Decisions, Markets, and Policies

MIT graduate course on energy economics covering market design, regulation, and policy analysis

MIT OCW: Engineering, Economics, and Regulation of the Electric Power Sector

MIT course bridging power systems engineering with electricity market economics and regulatory policy

Stephen Maher: Optimisation in the Real World

University of Exeter researcher applying OR to renewable energy, vaccine distribution logistics, and carbon-neutral supply chains. Creative applications including beer brewing optimization.

Coaching Actuaries

Premium exam preparation platform for SOA and CAS actuarial exams with practice problems, video lessons, and adaptive learning. Industry standard for exam preparation.

Great Learning Insurance Analytics Course

Free course covering insurance analytics fundamentals including customer segmentation, claims prediction, and fraud detection with Python implementations.

Freakonometrics Blog

Arthur Charpentier's blog covering actuarial science, machine learning, and R programming. Rich tutorials on insurance pricing, claims modeling, and statistical methods.

Swiss Association of Actuaries Tutorials

Professional tutorials on modern actuarial methods including machine learning for pricing, telematics, and reserving. Created by Mario Wuthrich and collaborators at ETH Zurich.

Modelling Extremal Events (Embrechts, Klüppelberg, Mikosch)

Comprehensive treatment of extreme value theory with 8,000+ citations. Covers applications in insurance and finance with rigorous mathematical foundations.

An Introduction to Statistical Learning of Extreme Values (Coles)

Foundational textbook on extreme value theory covering GEV and GPD distributions for tail modeling in climate, finance, and insurance.

Meta Engineering - Data Science

Large-scale experimentation, ML infrastructure, and data discovery at Facebook scale. Posts on causal inference and data tools.

Uber: Backtesting at Scale

Architecture for ~10 million backtests. Four backtesting vectors (cities, windows, variants, granularity). Go/Cadence workflows. Evolution from Omphalos framework to handle exponential growth.

Stripe: How We Built Radar

XGBoost→DNN migration, 85% training time reduction

eBay Tech Blog

eBay's engineering blog covering search ranking at scale in a marketplace with millions of listings, including relevance and seller quality signals.

Netflix: Predictive CPU Isolation of Containers

ML-based container isolation achieving 13% capacity reduction through predictive resource management.

Defense Acquisition University (DAU)

Official DoD training for acquisition professionals covering contracting, program management, and cost estimation

Machine Learning Specialization (Coursera)

Beginner-friendly 3-course series by Andrew Ng covering core ML methods (regression, classification, clustering, trees, NN) with hands-on projects.

Machine Learning for Economists (Hebrew University)

Complete course materials from Itamar Caspi and Ariel Mansura with R and Python tutorials on ML methods for economic research.

Open Source Economics: Structural Estimation

From UChicago's Masters in Computational Social Science. Structural vs. reduced-form, MLE, GMM, Simulated Method of Moments. Complete GitHub repositories with Python/Jupyter implementations.

Andrew Ng's Machine Learning Specialization

Comprehensive theoretical grounding redesigned 2022 with modern Python. Three-course sequence on supervised/unsupervised learning and recommender systems. 4.9/5 from 37,000+ reviews. Free to audit.

NLP for Economists (MGSE)

Munich Graduate School of Economics course on natural language processing methods for economics research by Sowmya Vajjala.

IMF Machine Learning for Economists Course

Course materials from Michal Andrle's IMF course on practical ML applications in economics and central banking.

Georgetown MA in Security Studies

Premier graduate program with Economics & Security requirement, training the next generation of defense analysts

QuantEcon Lectures

High-quality lecture series on quantitative economic modeling, computational tools, and economics using Python/Julia.

AEA Continuing Education: ML and Econometrics (2018)

9-part webcast series from Susan Athey and Guido Imbens on machine learning for economists. Freely available from the American Economic Association.

EconDL: Deep Learning for Economists

Companion website for Melissa Dell's JEL paper with demo notebooks, code examples, and tutorials on applying deep learning to economics research.

Coding for Economists (Arthur Turrell)

Python workflow for economists covering data transformation, econometrics, Bayesian inference, and ML. Modern Python-first approach.

Dario Sansone's ML Resources for Economists

Curated collection of machine learning resources specifically for economists including papers, code, and tutorials.

Hal Varian: Machine Learning and Econometrics (Berkeley)

Google's Chief Economist on bridging ML and econometrics, covering prediction vs inference, variable selection, and modern statistical approaches.

Anton Korinek Research

Korinek's research page with all papers, updates, and resources on AI and economics including the evolving JEL paper series.

Capitalisn't: Sendhil Mullainathan on Who Controls AI

Chicago Booth podcast exploring AI governance, algorithmic decision-making, and the economic implications of who shapes AI development.

Coding for Economists: NLP Chapter

Arthur Turrell's comprehensive guide covering text-as-data methods, topic modeling, and NLP applications for economists.

GenAI for Econ Substack

Anton Korinek's Substack newsletter with updates on LLM capabilities for economists and practical applications in research.

Knowledge Project #102: Sendhil Mullainathan

Deep conversation with Sendhil Mullainathan on behavioral economics, machine learning in social science, and decision-making under uncertainty.

QuantEcon DataScience

Economic modeling with data science from UBC Vancouver. Python-based applications in economics with real student project examples.

Brady Neal's Introduction to Causal Inference

14-week video course covering potential outcomes, DAGs, do-calculus, and causal discovery. Features guest lectures from Susan Athey, Alberto Abadie, and Yoshua Bengio. Bridges ML and econometric traditions.

Stanford ML & Causal Inference Short Course

Video lectures from Susan Athey, Jann Spiess, and Stefan Wager covering ML vs. econometrics, ATEs with propensity scores, CATE estimation with causal forests, and loss functions for causal inference.

QuantEcon: ML in Economics

Interactive Python tutorials on machine learning for prediction and causal inference from the QuantEcon project.

Meta: Instagram Notification Management with ML and Causal Inference

How Instagram uses ML and causal methods to optimize notification delivery, balancing engagement with user experience.

Airbnb: ACE - Artificial Counterfactual Estimation

Machine learning-based causal inference at Airbnb. Using ML to estimate counterfactuals when traditional experimental methods aren't feasible.

AEA/AFA 2019: Impact of Machine Learning on Economics

Susan Athey's joint luncheon address on how ML is reshaping economic research, prediction policy problems, and heterogeneous treatment effects.

Dario Sansone: ML for Economists Resources

Curated list of ML and causal inference resources for economists. Papers, tools, and comprehensive meta-resource collection.

AEA: Machine Learning and Econometrics (Athey/Imbens)

9 hours from two of the most influential computational economists. ML vs. causal inference, heterogeneous treatment effects, LASSO/random forests, causal forests, policy learning. Athey pioneered ML in economics; Imbens won 2021 Nobel.

EconTalk: Susan Athey on ML, Big Data, and Causation

Susan Athey discusses how machine learning transforms economic research, the importance of big data for causal inference, and bridging CS/economics methodologies.

Stanford Fintech: Sendhil Mullainathan on ML as Tool for Science

ABFR webinar where Mullainathan discusses ML's role in scientific discovery, moving beyond prediction to understanding causal mechanisms.

Kaggle's Intermediate Machine Learning

Hands-on XGBoost with graded exercises. Covers missing values, categorical encoding, pipelines, cross-validation, then XGBoost tuning (n_estimators, early_stopping, learning_rate). Free certificate.

Analytics Vidhya: Hyperparameter Tuning Guide

Systematic tuning methodology from Kaggle winners. Sequential approach: fix tree params, tune learning rate/iterations, add regularization. Key insight: 10× decrease in learning_rate needs ~10× increase in n_estimators.

Deep Learning for Single-Cell Analysis (scverse tutorials)

Tutorials on deep generative models for single-cell genomics. Covers scVI, totalVI, and related ZINB autoencoder approaches for sparse count data.

Conformal Prediction Tutorial (Angelopoulos & Bates)

Comprehensive tutorial on conformal prediction providing distribution-free prediction intervals with finite-sample guarantees. Works as a post-hoc wrapper on any model including neural networks.

Neural Survival Analysis with pycox

Documentation and tutorials for PyTorch survival analysis. Implements DeepSurv, DeepHit, Cox-Time with neural network architectures.

Instacart: Predicting Availability of 200M Grocery Items

XGBoost with 130 features scoring 200M+ items every 60 minutes. 15x items with 1/5 resources.

SDV Getting Started Guide

Official Synthetic Data Vault documentation covering GaussianCopula, CTGAN, and TVAE models for tabular data synthesis.

Neptune.ai: When to Choose CatBoost Over XGBoost

Algorithm selection with benchmark comparisons. Explains CatBoost's ordered boosting (preventing target leakage), symmetric vs. asymmetric trees. Decision framework practitioners need.

MIT OCW: Dynamic Programming (Bertsekas)

6 advanced lectures (~12 hours) from the definitive DP authority. Approximate DP, large-scale infinite horizon problems, policy iteration with function approximation, temporal difference, neuro-dynamic programming.

IMF: Cross-Validation for Economists

IMF training material on applying cross-validation techniques in economic research, bridging ML best practices with econometric applications.

MLJAR: Feature Importance with XGBoost

Definitive guide covering three importance methods: gain, weight, and SHAP. Complete Colab code comparing built-in importance vs. permutation vs. SHAP values. Essential for model interpretation.

UChicago Law: Discrimination by Algorithm and People

Sendhil Mullainathan examines algorithmic discrimination, comparing ML-based decisions to human decisions, with policy implications.

Interpretable Machine Learning Book (Christoph Molnar)

The definitive online book on ML interpretability: SHAP, LIME, PDP, feature importance. Essential for understanding black-box model predictions.

Google Machine Learning Crash Course

15-hour interactive course originally for Google engineers, refreshed 2024 with LLMs/AutoML. Covers supervised learning, feature engineering, and production ML with Colab exercises. Teaches exact mental models Google engineers use.

Beyond Jupyter

Software design principles for ML applications. Go from messy notebooks to maintainable, modular code with OOP essentials and refactoring guides.

Beyond Jupyter (TransferLab)

Teaches software design principles for ML—modularity, abstraction, and reproducibility—going beyond ad hoc Jupyter workflows. Focus on maintainable, production-quality ML code.

Gymnasium Documentation

Official Farama Foundation documentation for Gymnasium RL environments including tutorials on building custom environments.

Richard Oberdieck: Modern OR Software Engineering

Modern software engineering practices for optimization. Includes 'LLM-ify me - Optimization edition' exploring AI-OR integration and Python modeling patterns.

Made With ML

Implementation-first approach: build models from scratch with NumPy before PyTorch. Emphasizes clean, production-quality code with proper software engineering practices. By Goku Mohandas (ex-Apple ML).

DoorDash: Using ML and Optimization to Solve Dispatch

DeepRed engine combining ML prediction layer with MIP optimization for batching decisions.

Eugene Yan: Bandits for Recommender Systems

The definitive practitioner's guide synthesizing implementations from 12+ tech companies (Spotify, Netflix, Yahoo, DoorDash, Twitter, Alibaba, Amazon). Covers ε-greedy, UCB, Thompson Sampling.

Eppo: How Netflix, Lyft, and Yahoo Use Contextual Bandits

Case studies: Netflix artwork personalization, Lyft pricing optimization, Yahoo news with LinUCB. Explains why contextual bandits beat full recommenders for smaller action spaces.

Fast.ai Practical Deep Learning for Coders

Top-down approach: deploying models by lesson 2, then progressively revealing mechanics. Part 1: vision, NLP, tabular, collaborative filtering. Part 2: backprop to Stable Diffusion. Alumni at Google Brain, OpenAI, Tesla.

Eugene Yan: System Design for Recommendations

Production patterns from Alibaba, Facebook, DoorDash, LinkedIn in a 2×2 framework (offline/online × retrieval/ranking). By Amazon Principal Applied Scientist. Referenced by NVIDIA as canonical industry reading.

Netflix Technology Blog: Recommendation Systems

How Netflix Prize pioneers continue innovating. Foundation models with transformers, multi-task learning across surfaces, RecSysOps for production monitoring at 200M+ user scale. Lessons unavailable elsewhere.

Andrew Ng's Machine Learning Specialization

Comprehensive theoretical grounding redesigned 2022 with modern Python. Three-course sequence on supervised/unsupervised learning and recommender systems. 4.9/5 from 37,000+ reviews. Free to audit.

Airbnb: Listing Embeddings for Similar Listing Recommendations

Word2Vec-inspired embeddings from 800M+ search sessions achieving 21% CTR increase.

Stitch Fix Algorithms Blog

Demand forecasting, inventory optimization, and personalization. Unique blend of fashion retail + serious data science.

Google's Recommendation Systems Course

Industry-standard architecture: candidate retrieval → scoring → re-ranking. Built by YouTube RecSys engineers. 4-hour course on collaborative filtering, matrix factorization, embeddings, deep approaches. YouTube case study at 2B+ user scale.

TensorFlow Recommenders Tutorials

Executable code for two-tower architecture used at Google, YouTube, Pinterest. MovieLens examples: user/item embeddings, retrieval models, ranking layers, serving with approximate nearest neighbors. Concept to deployment.

Jay Alammar's Illustrated Transformer

Definitive visual guide to attention mechanisms, referenced at Stanford, Harvard, MIT, Princeton, CMU. Step-by-step illustrations of self-attention, multi-head attention, positional encoding. Covers BERT, GPT-2, retrieval transformers.

Anthropic's Prompt Engineering Tutorial

Definitive prompting from Claude's creators. 26,000+ GitHub stars. Interactive notebooks on direct prompting, chain-of-thought, output formatting, hallucination avoidance, tool use. 'Best LLM vendor documentation' - Simon Willison.

Google NotebookLM

AI-powered research notebook that auto-generates podcasts and summaries from uploaded research papers and documents.

Ahead of AI (Sebastian Raschka)

ML & AI research newsletter from Sebastian Raschka. Deep technical coverage of LLMs, model architectures, training techniques, and AI trends. Author of 'Build a Large Language Model From Scratch'.

Eugene Yan: Patterns for Building LLM-based Systems

7 production patterns: Evals, RAG, Fine-tuning, Caching, Guardrails, Defensive UX, User Feedback. 66-minute read with evaluation metrics (BLEU, ROUGE, BERTScore), RAG patterns, fine-tuning decisions. From Amazon experience.

LangChain Academy: Intro to LangGraph

Most comprehensive free agent-building course. 6-hour, 55-lesson course on state management, memory, human-in-the-loop, parallelization, deployment. Used in production at Klarna, LinkedIn, Elastic.

FreeCodeCamp: RAG from Scratch

Deep RAG understanding by Lance Martin (LangChain engineer, Stanford PhD). 2.5-hour video on advanced techniques: Multi-Query, RAG Fusion, Decomposition, Step Back, HyDE, Corrective RAG, Self-RAG patterns.

DeepLearning.AI Short Courses

Rapid skill-building in 1-2 hours. Key free courses: LangChain for LLM Apps (Harrison Chase), Building Systems with ChatGPT, Functions/Tools/Agents. Interactive Jupyter notebooks, zero setup.

Anthropic Prompt Engineering Documentation

Systematic approach to prompting Claude models effectively. Official documentation with best practices and examples.

DAIR.AI Prompt Engineering Guide

Industry-standard open-source guide covering all prompting techniques for LLMs. Supports 13 languages with comprehensive coverage of chain-of-thought, few-shot, and advanced prompting methods.

Temporal Fusion Transformer: Complete Tutorial

End-to-end TFT with PyTorch Forecasting. Handles heterogeneous features (static, time-varying known/unknown). Interpretability via variable importance and attention. Shows when TFT outperforms simpler methods.

DoorDash: ELITE Ensemble Learning

ELITE (Ensemble Learning for Improved Time-series Estimation). Addresses accuracy vs. speed/cost tradeoffs. Scales to tens of thousands of targets. Practical engineering decisions for when perfect is enemy of good.

Instacart Anytime: Data Science Paradigm

End-to-end system: forecasting integrates with supply planning and capacity decisions. Key metrics: Availability, Idleness, Unmet Demand. Multi-horizon forecasting (weeks ahead for acquisition, hourly for store-level).

Uber: Forecasting Introduction

Written by M4 Competition winner team. Covers 15 million trips/day across 600+ cities. Explicitly addresses ML vs. statistical methods decision. Use cases: marketplace, capacity planning, marketing.

Time Series Forecasting with Lag-Llama

Foundation models landscape (Lag-Llama, TimesFM, Moirai, TimeGPT-1). Zero-shot vs. fine-tuning decision framework. Probabilistic forecasts with uncertainty quantification. Complete Python with GluonTS.

Time Series Handbook: LightGBM for M5

Complete Jupyter Book with runnable code. LightGBM MAE (200.5) vs. naive baseline (698.0). Feature engineering (lags, rolling windows), recursive vs. direct forecasting, hyperparameter tuning. Free via GitHub with Binder.

Blocked Time Series Cross Validation

Addresses critical issue: expanding window CV produces overly optimistic estimates. Drop-in sklearn-compatible code. Explains why blocked CV gives realistic production performance estimates.

TensorFlow Time Series Forecasting Tutorial

Official Google documentation with production-quality code. Builds models incrementally: linear → dense → CNN → LSTM. Includes baseline comparisons so you can assess if DL is worth the complexity. Runnable in Colab.

M5 Competition Analysis: Learnings and Winning Solutions

Synthesizes learnings from 5,558 teams on 42,840 time series. Key finding: ML beats statistical when you have many correlated series, exogenous variables, hierarchical structure. LightGBM vs. N-BEATS vs. seq2seq comparison.

Conformal Prediction Intervals for Time Series

Distribution-free uncertainty quantification without Gaussian assumptions. Model-agnostic approach works with any forecasting method. Addresses limitation of bootstrap (only captures data uncertainty). MAPIE implementation.

OpenAI Cookbook: Semantic Search with Embeddings

Modern embedding-based retrieval end-to-end. Embedding generation with OpenAI API, Pinecone vector database, cosine similarity search. Foundation for semantic search and RAG systems.

OLX Engineering: From RankNet to LambdaMART

Clearest learning-to-rank explanation with code. Why ranking differs from classification, pointwise vs. pairwise vs. listwise approaches. Implementing RankNet and LambdaMART with XGBoost rank:pairwise and rank:ndcg.

Airbnb: ML-Powered Search Ranking

Masterclass in production search evolution. 4-stage journey from baseline to personalized GBDT ranking with A/B test results (+13%, +7.9%, +5.1% booking improvements). Feature engineering for two-sided marketplaces.

Eugene Yan: Position Bias in Search

Measurement and mitigation techniques: RandPair, FairPairs, propensity scoring. Essential for production ranking systems where position corrupts training data.

LinkedIn: AI Behind Recruiter Search

Enterprise-scale search: multi-layer ranking (L1 retrieval → L2 ranking), evolution from linear to GBDT to neural, GLMix personalization. Among the largest learning-to-rank systems in production.

LinkedIn Engineering

Professional network data science, feed ranking, economic graph insights. ML and economics at scale.

Netflix Tech Blog

Streaming personalization, A/B testing at scale, recommendations. How Netflix builds data products for 200M+ subscribers.

Etsy: Building Marketplace Search and Personalization

Etsy engineering on building search for a handmade goods marketplace. Covers ranking, personalization, and balancing buyer and seller interests.

PyMC-Marketing CLV Quickstart

CLV basics, RFM analysis, BG/NBD models — free official docs

PyMC-Marketing Documentation

BG/NBD and Gamma-Gamma CLV tutorials

Bruce Hardie's CLV Papers

Mathematical foundations of CLV models

Growth Accounting & Backtraced Growth Accounting

Standard framework for user lifecycle states (New, Retained, Churned, Stale, Resurrected) with weighted backtrace views

Juan Orduz: Bayesian Marketing Methods

Principal Data Scientist at PyMC Labs with PhD in Mathematics. 50+ deep technical posts on media effect estimation, adstock/saturation curves, CLV modeling, and synthetic controls.

The Power User Curve (a16z)

The L30/L28 framework coined by Facebook's growth team. Why Power User Curves beat DAU/MAU: reveals variance, identifies power users, customizable for core actions. Used by a16z to evaluate startups.

Airbnb: Measuring Listing Lifetime Value

Production function approach modeling incrementality based on supply-demand balance. How Airbnb values new listings in their marketplace.

Wharton Customer Analytics (Coursera)

The gold-standard course on customer analytics from Wharton's Customer Analytics Initiative. Taught by Eric Bradlow, Peter Fader, and Raghuram Iyengar covering CLV, segmentation, and predictive analytics.

Bruce Hardie's BTYD Tutorials

Step-by-step mathematical derivations of Pareto/NBD, BG/NBD, and other BTYD models from one of the field's pioneers. Essential reference for implementing CLV models from scratch.

Wharton Customer Analytics Initiative (WCAI)

World's preeminent customer analytics research center. Pioneered industry-academic collaboration with access to proprietary datasets and practitioner-focused research.

How to Measure Cohort Retention (Lenny's Newsletter)

The most comprehensive retention measurement guide. SQL implementations, bounded vs unbounded retention definitions, visualization best practices. When to use X-day vs unbounded retention.

Lenny's Newsletter: How Duolingo Reignited User Growth

Case study on gamification, streaks, and retention mechanics that drove 4.5x growth

A Quantitative Approach to Product-Market Fit (Tribe Capital)

The foundational text on growth accounting. MAU growth accounting AND revenue growth accounting. Quick Ratio, Gross Retention, Net Churn explained by the team that pioneered it.

Ultimate Guide: Activation (Aakash Gupta)

Traces activation history from Facebook's 2008 growth team, including Chamath's '7 friends in 10 days' discovery. The Setup → Aha → Habit framework with data-backed examples.

Slack's 2000 Messages Activation Metric

Documents Slack's activation discovery — after 2,000 messages sent per team, 93% remain active. How they identified this leading indicator. Conversion rate significantly above 5% SaaS average.

RevenueCat Sub Club: Subscription Analytics

Definitive resource for mobile subscription analytics. Bi-weekly newsletter and podcast featuring practitioners from Duolingo, Strava, and Lose It! with actual retention data.

Eva Ascarza: Retention Futility Research

Harvard Business School professor challenging conventional retention management. Her paper 'Retention Futility' demonstrates that standard churn interventions may backfire.

Juan Orduz: Bayesian Marketing Methods

Principal Data Scientist at PyMC Labs with PhD in Mathematics. 50+ deep technical posts on media effect estimation, adstock/saturation curves, CLV modeling, and synthetic controls.

Google Analytics for Marketing

Free analytics for marketing — official Google tutorials

Recast Blog: MMM Verification

Michael Kaminsky (former Director of Analytics at Harry's) on MMM verification, hypothesis testing, model falsifiability, and when MMM investment makes sense.

Mario Filho: Forecastegy

Kaggle Competitions Grandmaster (#12 globally) and former Lead Data Scientist at Upwork. Hands-on MMM implementation tutorials with real advertising data.

Mike Taylor: Vexpower MMM Tutorials

Former Ladder.io co-founder who managed $50M+ in marketing spend across 8,000 experiments. Most accessible MMM tutorials for LightweightMMM, Robyn, and Uber Orbit.

Marketing Science Institute (MSI)

Bridge between marketing academia and industry. Sets annual research priorities and publishes working papers on topics from brand measurement to customer analytics.

Byron Sharp: How Brands Grow

Ehrenberg-Bass Institute director and leading critic of marketing pseudoscience. Established empirical laws (Double Jeopardy, Duplication of Purchase) challenging myths about brand loyalty.

021 Newsletter: Marketing & Data Teams Bridge

Barbara Galiza bridging marketers and data teams (7,000+ subscribers). When to use click attribution vs MMM, how to structure incrementality testing, marketing data infrastructure.

Kevin Simler: Ads Don't Work That Way

Essential advertising theory essay distinguishing cultural imprinting from emotional inception. Explains why broadcast advertising works differently than targeted digital.

AdKDD Workshop Papers

Applied research from Google, Meta, Amazon

Lumen Research: Attention Metrics

Leading research on attention metrics as viewability's evolution. Research shows attention is 3x better at predicting outcomes than viewability.

AppsFlyer Privacy Sandbox Hub

Comprehensive resource comparing iOS and Android privacy frameworks. Essential for understanding cross-platform privacy measurement approaches.

Branch Resources: Privacy-Centric Measurement

Deep linking and mobile attribution provider with excellent content on making sense of aggregate data and privacy-centric measurement approaches.

Mobile Dev Memo: Post-ATT Marketing Measurement

Eric Seufert's definitive voice on mobile marketing measurement. Weekly deep-dives on SKAdNetwork, iOS attribution challenges, and econometric marketing measurement.

Remerge Findings: Incrementality Testing Approaches

Technical breakdowns of incrementality testing methods from a DSP perspective. Covers intent-to-treat, PSA, ghost ads, and ghost bids with clear pros and cons.

Adjust Blog: Mobile Attribution & Privacy

Leading mobile measurement partner with current implementation guidance for SKAdNetwork, AdAttributionKit, and Privacy Sandbox.

Media Rating Council (MRC)

Industry body setting viewability standards and measurement accreditation for digital advertising

Tallys Yunes: OR by the Beach

Associate Professor at University of Miami focusing on making optimization accessible. Downloadable 'Optimization Games for the Young' and everyday optimization examples.

Ryan O'Neil: Real-Time Optimization

Co-founder of Nextmv, PhD from George Mason under Karla Hoffman. Writes about real-time optimization for delivery platforms, hybrid optimization and decision diagrams.

QuantEcon: Linear Programming Introduction

Python notebooks from Nobel Laureate Sargent. LP with economics applications using SciPy and OR-Tools.

Paul Rubin: OR in an OB World

Professor Emeritus at Michigan State with 33+ years experience. Most technically detailed academic blog with specific CPLEX tips, Java/R code snippets, and reader Q&A.

SolverMax: Python OR Library Comparison

13-article series comparing Python OR libraries plus comprehensive directory of optimization blogs with summaries and notable posts.

Franco Peschiera: PuLP Maintainer

PuLP library maintainer publishing 'PuLP: past, present and future' and Timefold integration posts. Insider perspective on open-source OR library development.

MIT 6.046J Lecture 15: Linear Programming

Video intro from algorithmic perspective. LP formulation, reductions, and simplex method.

GILP: Geometric Interpretation of Linear Programs (Cornell)

Academic-grade visualization (ACM SIGCSE 2023). Shows feasible regions, simplex iterations, branch-and-bound.

Dominik Krupke: CP-SAT Primer

The most comprehensive unofficial guide to Google OR-Tools' CP-SAT solver. Chapters cover modeling patterns, parameter tuning, benchmarking methodology, and large neighborhood search.

Nathan Brixius: ML + Optimization

Former Microsoft Solver Foundation developer bridging optimization and machine learning. Posts on chaining ML and optimization models, solving historical IP problems with modern solvers.

Alain Chabrier: Column Generation with CPLEX

Former IBM Decision Optimization Senior Technical Staff Member. Authoritative content on column generation with docplex/CPLEX. His PhD solved 17 previously open Solomon VRP benchmark instances.

NSPLib: Nurse Scheduling Benchmarks (Ghent University)

Benchmark instances for nurse scheduling with downloadable datasets and solutions. Covers genetic algorithms, scatter search, and nurse rerostering.

PuLP Official Documentation

Complete LP/MIP documentation with case studies: blending problem, Sudoku, transportation. Multiple solver support.

DoorDash: Next-Generation Dasher Dispatch Optimization

Rare solver benchmarking transparency — compares CBC, XPress, CPLEX, Gurobi (34x faster than CBC).

Instacart: Delivering Optimal Shopping Experiences (Gurobi)

Why Instacart chose commercial solvers. Reliability and innovation speed from Gurobi.

Discrete Optimization (Coursera)

Van Hentenryck's course — actually makes you implement

Google OR-Tools Python Guide

Official documentation with setup and examples. CP-SAT solver won MiniZinc Challenge 2013-2024.

DoorDash: Next-Generation Optimization for Dasher Dispatch

Migration to MIP with Gurobi achieving 34x faster optimization than CBC.

Google OR-Tools: VRP + VRPTW Tutorial

Core logistics vocabulary (depot, fleet, constraints) with working Python baseline

Real Python: Linear Programming with Python

Comprehensive tutorial covering visualization, feasible regions, SciPy, PuLP, and mixed-integer programming.

Austin Buchanan: Farkas' Dilemma

Oklahoma State Associate Professor publishing accessible tutorials on Benders decomposition, Lagrangian techniques for k-median, and political redistricting applications.

Hands-On Mathematical Optimization with Python (MO-book)

50+ Jupyter notebooks from Postek (BCG), Zocca, Gromicho (ORTEC), and Kantor (Notre Dame). Linear optimization through optimization under uncertainty with Pyomo implementations.

Jeffrey Kantor: Pyomo Cookbook

381+ GitHub stars. Practical Pyomo modeling examples that complement official documentation. From Notre Dame professor.

Richard Oberdieck: Modern OR Software Engineering

Modern software engineering practices for optimization. Includes 'LLM-ify me - Optimization edition' exploring AI-OR integration and Python modeling patterns.

Timefold Blog

Founded by OptaPlanner creator Geoffrey De Smet (17+ years OR experience). Employee rostering, nurse scheduling, and constraint programming with Java/Kotlin/Python.

Erwin Kalvelagen: Yet Another Math Programming Consultant

Decades of practical modeling wisdom from a GAMS/AMPL/CPLEX consultant. Large sparse transportation models, MINLP formulations, solver tuning tricks, and creative problems like Wordle optimization.

Modeling Discrete Optimization (Coursera)

University of Melbourne's course on constraint programming, local search, and MIP. Covers MiniZinc modeling language.

Georgia Tech (Ratliff): 10 Rules for Supply Chain Optimization

Practitioner checklist for scoping, data readiness, constraints, deployment — free PDF

Kevin Gue: Warehouse Design

Senior Director of R&D at Fortna, formerly academia at Naval Postgraduate School, Auburn, and Louisville. DC Velocity columns on warehouse design and order picking routing.

Google OR-Tools: VRP + VRPTW Tutorial

Core logistics vocabulary (depot, fleet, constraints) with working Python baseline

Feasible Newsletter

Weekly OR industry news by Borja Menendez. Real-world case studies from UPS ORION and Walmart Route Optimization, solver announcements, and career guidance.

Nextmv Blog

From former Convoy and Uber operations researchers. Bridges open-source tools and production systems. DecisionFest recordings feature IKEA, Walmart, Carvana, and Toyota.

Alain Chabrier: Column Generation with CPLEX

Former IBM Decision Optimization Senior Technical Staff Member. Authoritative content on column generation with docplex/CPLEX. His PhD solved 17 previously open Solomon VRP benchmark instances.

VRPSolver: Column Generation for Vehicle Routing

Cutting-edge branch-cut-and-price algorithms by Eduardo Uchoa, Artur Pessoa, and Lorenza Moreno. State-of-the-art academic work with production solver implications.

Stephen Maher: Optimisation in the Real World

University of Exeter researcher applying OR to renewable energy, vaccine distribution logistics, and carbon-neutral supply chains. Creative applications including beer brewing optimization.

Stitch Fix: Algorithms Tour

The single best piece of data journalism in tech. Interactive, animated tour of how they combine styles, logistics, and feedback loops.

Amazon Science

Research from Amazon's scientists. Causal inference, supply chain optimization, pricing, and forecasting.

Afi Labs: Ride-Share Dispatch Algorithms

Complete worked examples for ride-share dispatch with full code. Explains why greedy nearest-driver matching fails compared to optimal trip chaining.

DoorDash Engineering

marketplace analytics, delivery optimization, and experimentation. Great posts on real-time pricing and logistics.

DoorDash: ML + Optimization for Dispatch

Clearest 'real system' explanation: predictions feed optimizer, then simulation closes the loop

Amazon Science: Operations Research and Optimization

Portal to Amazon's OR research on inventory planning, last-mile delivery, and fulfillment at massive scale.

Ryan O'Neil: Real-Time Optimization

Co-founder of Nextmv, PhD from George Mason under Karla Hoffman. Writes about real-time optimization for delivery platforms, hybrid optimization and decision diagrams.

DoorDash: Next-Generation Dasher Dispatch Optimization

Rare solver benchmarking transparency — compares CBC, XPress, CPLEX, Gurobi (34x faster than CBC).

Laura Albert: Punk Rock Operations Research

2023 INFORMS President, NSF CAREER Award winner. Most-read academic OR blog with 84,600+ annual hits. Emergency response optimization, ambulance dispatch, homeland security analytics.

DoorDash: The Dasher Dispatch System

Technical deep dive into how DoorDash assigns deliveries to Dashers. Covers matching algorithms, optimization objectives, and real-time constraints.

Lyft: Solving Dispatch in a Ridesharing Problem Space

Hungarian algorithm and LP relaxation for real-time bipartite matching. Technical deep-dive on driver-rider matching optimization.

DoorDash: Using ML and Optimization to Solve Dispatch

DeepRed engine combining ML prediction layer with MIP optimization for batching decisions.

Erlang C Calculator

Interactive online calculator for M/M/c queue metrics: service level, delay probability, average waiting time.

MIT 6.262 Discrete Stochastic Processes

MIT OpenCourseWare covering Poisson processes, Markov chains, renewal theory, and queueing applications. Complete lecture videos and problem sets.

edX Queuing Theory: from Markov Chains to Multi-Server Systems

IMT course covering M/M/1, Erlang formulas, with Python labs. Self-paced online learning.

MIT 15.070J Advanced Stochastic Processes

Graduate-level MIT course on heavy traffic theory for queueing systems. Advanced mathematical treatment.

AWS Builders Library: Avoiding Insurmountable Queue Backlogs

AWS best practices for queue management, backpressure, and avoiding cascading failures.

Shopify: Capacity Planning at Scale

Black Friday/Cyber Monday planning with GCP traffic scenarios at massive scale.

Erlang A Calculator

M/M/c+M model calculator with abandonments using Garnett-Mandelbaum-Reiman approximations.

Kendall Notation Tutorial

Interactive tutorial on the A/B/C/K/N/D queueing notation system introduced by David Kendall in 1953.

Kingman's Formula Tutorial

Practical guide to the VUT approximation for G/G/1 waiting time. Essential for heavy traffic analysis.

Little's Law 50th Anniversary Paper

Retrospective on L = λW proving average customers equals arrival rate times average time, regardless of distributions.

MIT 15.053: Optimization Methods in Management Science

Undergraduate course on LP with geometry and visualization before algebra. Interactive spreadsheet exercises.

SciPy Lecture Notes: Mathematical Optimization

Academic tutorial with visual explanations. Gradient descent, BFGS, Nelder-Mead with convergence visualizations.

Convex Optimization (Boyd & Vandenberghe)

The bible of convex optimization — free online, universally cited. Covers LP, QP, SDP, and more.

Stanford EE364A (YouTube)

Boyd's legendary lectures on convex optimization. The gold standard for learning optimization theory.

Modeling Discrete Optimization (Coursera)

University of Melbourne's course on constraint programming, local search, and MIP. Covers MiniZinc modeling language.

Nathan Brixius: ML + Optimization

Former Microsoft Solver Foundation developer bridging optimization and machine learning. Posts on chaining ML and optimization models, solving historical IP problems with modern solvers.

CVXPY Short Course

Hands-on convex optimization in Python. Learn to model and solve real problems with CVXPY.

SolverMax: Python OR Library Comparison

13-article series comparing Python OR libraries plus comprehensive directory of optimization blogs with summaries and notable posts.

Tallys Yunes: OR by the Beach

Associate Professor at University of Miami focusing on making optimization accessible. Downloadable 'Optimization Games for the Young' and everyday optimization examples.

awesome-optimization

Curated list of courses, books, libraries, and frameworks across convex optimization, discrete optimization, and metaheuristics. Comprehensive starting point regularly updated.

Fiverr Engineering Blog

How Fiverr structures their freelance marketplace, from gig discovery to pricing and seller success mechanics.

Hagiu: Multi-Sided Platforms - From Microfoundations to Design

Academic treatment of platform design principles. Bridges economic theory with practical platform design decisions.

Hagiu & Wright: When to Open a Platform (HBR)

HBR analysis of when platforms should allow third-party developers. Framework for deciding between closed and open platform strategies.

Jeff Jordan (a16z): Marketplace 100

Annual ranking and analysis of the largest consumer marketplaces. Framework for understanding marketplace categories and business models.

Platform Papers (Joost Rietveld)

Academic platform research translated for practitioners. Bridges academic literature with practical platform strategy.

Andrei Hagiu: Strategic Decisions for Multisided Platforms

MIT Sloan Management Review guide to key strategic choices for platform businesses: sides to bring on board, design, and pricing decisions.

Platform Revolution Book Overview

Overview of the seminal book on platform business models. Parker, Van Alstyne, and Choudary's framework for understanding platform dynamics.

Parker, Van Alstyne & Choudary: Pipelines, Platforms, and the New Rules of Strategy

HBR article summarizing Platform Revolution. Contrasts pipeline (linear) businesses with platform businesses and their different strategic imperatives.

Instacart: Building for Balance (SAGE v2)

Unique four-sided marketplace perspective (consumers, shoppers, retailers, brands). How Instacart balances all sides of their complex marketplace.

Bill Gurley: All Markets Are Not Created Equal

Essential framework for evaluating marketplace businesses. Gurley identifies 10 factors that distinguish great marketplaces from mediocre ones, including fragmentation, frequency, and payment facilitation.

Jean Tirole: Two-Sided Markets - A Progress Report

Nobel laureate's comprehensive survey of two-sided market theory. Covers pricing, platform competition, and welfare implications of multi-sided platforms.

Rochet & Tirole: Platform Competition in Two-Sided Markets

Foundational academic paper on platform economics. Develops theory of pricing and competition when platforms must attract multiple user groups.

EconTalk: Hal Varian on Technology

Wide-ranging discussion with Google's Chief Economist on technology adoption, internet economics, and data-driven decision making.

DoorDash: Building a Successful Three-Sided Marketplace

DoorDash engineering explains the unique challenges of balancing three sides: merchants, dashers, and consumers in their delivery marketplace.

Lyft: Simulating a Ridesharing Marketplace

Lyft engineering blog on counterfactual simulation framework for rideshare marketplace optimization.

Eugene Wei: Seeing Like an Algorithm (TikTok)

Deep analysis of TikTok's success through algorithmic content discovery. Explains why TikTok's approach differs from social graph-based networks.

Uber: Unleashing the Power of Ads Simulation

Uber Eats engineering post on building an ads marketplace simulator for testing ad ranking and bidding strategies.

Amazon Science: How Amazon Robots Navigate Congestion

Algorithms computing social rules for 8,000+ robots per fulfillment center.

Bill Gurley: How to Miss By a Mile (Uber TAM)

Analysis of TAM (Total Addressable Market) estimation errors. Explains why most TAM analyses are flawed and how to think about market sizing for tech companies.

Airbnb: Measuring Listing Lifetime Value

Production function approach modeling incrementality based on supply-demand balance. How Airbnb values new listings in their marketplace.

Lyft: Quantifying Efficiency in Ridesharing

Efficiency isn't speed—it's an economic equilibrium. A masterclass in defining the objective function for marketplace optimization.

Tim Roughgarden's CS269I: Incentives in Computer Science

20+ hours of video with publication-quality notes. Covers Gale-Shapley, NRMP matching, deferred acceptance, strategyproofness proofs, cryptocurrency incentives. Uniquely bridges classical stable matching with modern applications.

Uber: Reinforcement Learning for Marketplace Balance

Largest RL deployment for matching in ridesharing—400+ cities globally. How Uber uses reinforcement learning to balance supply and demand in real-time.

Upwork Engineering Blog

Upwork's approach to matching freelancers with projects. Covers ML-based matching, skill inference, and marketplace quality.

Etsy: Building Marketplace Search and Personalization

Etsy engineering on building search for a handmade goods marketplace. Covers ranking, personalization, and balancing buyer and seller interests.

LinkedIn Engineering: Marketplace Optimization

How LinkedIn optimizes their talent marketplace to match candidates with opportunities while balancing multiple stakeholder interests.

DoorDash: The Dasher Dispatch System

Technical deep dive into how DoorDash assigns deliveries to Dashers. Covers matching algorithms, optimization objectives, and real-time constraints.

Tim Roughgarden's CS364A: Kidney Exchange

Definitive algorithmic treatment of kidney exchange. Covers Top Trading Cycles, cycle packing, incentive-compatible organ allocation. The actual algorithms used by the Alliance for Paired Kidney Donation.

Thumbtack Engineering Blog

How Thumbtack connects customers with local service professionals. Covers matching, pricing, and marketplace dynamics.

Airbnb Engineering: Two-Sided Marketplace Matching

Unique focus on 'both sides must accept' constraint. Host acceptance prediction, listing embeddings, cold start solutions. Shows how to infer host preferences from behavior—3.75% booking improvement.

Instacart: No Order Left Behind; No Shopper Left Idle

Monte Carlo simulations balancing supply/demand with Markov models for marketplace matching.

Lyft: Dynamic Pricing to Sustain Marketplace Balance

Evolution of Lyft's PrimeTime surge algorithm. Explains undersupply spirals and iterative fixes for two-sided marketplace pricing.

Airbnb: Learning Market Dynamics for Optimal Pricing

Airbnb Engineering post combining ML and structural modeling for Smart Pricing. Shows simulation-based approach to pricing.

Lyft Engineering

Rideshare economics, forecasting, and marketplace efficiency. Technical deep-dives on pricing, dispatch, and causal inference.

Rochet & Tirole: Two-Sided Markets (RAND)

Seminal paper introducing the economics of two-sided markets. Analyzes how platforms set prices when they serve distinct but interdependent customer groups.

Bill Gurley: A Rake Too Far

Classic analysis of take rates in marketplaces. Explains why high take rates invite competition and examines optimal pricing strategies for platform businesses.

Uber Engineering

Surge pricing, marketplace design, causal inference at scale. See how researchers tackle real problems at Uber.

CEPR VoxEU: Doing Economics at Google

Hal Varian discusses the practice of economics in a tech company: auction design, pricing, and empirical work at massive scale.

Uber: Simulated Marketplace with ML

Agent-based discrete event simulation for testing dispatch algorithms safely. How Uber builds digital twins of their marketplace to test pricing and matching changes.

Uber: Driver Surge Pricing

Shows why multiplicative surge is NOT incentive-compatible; presents the additive driver surge mechanism now in production. Foundational work on incentive design for gig economy platforms.

LinkedIn: The Economic Graph

LinkedIn's vision for mapping the global economy. How they use data to understand labor markets, skills, and economic opportunity.

Simon Rothman (a16z): How to Build a Marketplace

Former eBay Motors GM and a16z partner's comprehensive guide to marketplace building. Covers liquidity, matching, and scaling strategies.

Instacart Tech Blog

Marketplace balancing, delivery optimization, demand forecasting. Making on-demand grocery profitable.

The Cold Start Problem (Andrew Chen)

Atomic Networks and tipping points of two-sided marketplaces — why growth stalls

DoorDash: Real-Time Optimization of Delivery Operations

How DoorDash optimizes delivery operations in real-time, balancing Dasher earnings, merchant experience, and consumer wait times.

Eugene Wei: Invisible Asymptotes

Former Amazon exec explains how to identify hidden growth ceilings. Uses Amazon examples to show how companies can spot and overcome invisible constraints.

Dean Eckles: Blog on Network Experiments and Social Influence

MIT professor and former Facebook data scientist. Deep expertise in network experiments, social influence, and randomization at Facebook scale.

Eugene Wei: Status as a Service (StaaS)

Landmark essay analyzing social networks through the lens of status. Explains why networks rise and fall based on their ability to provide status games.

Meta: How Meta Tests Products with Strong Network Effects

Cluster experiments, power vs purity tradeoffs. How Facebook handles experimentation when treatment effects spill over between users.

a16z: Measuring Network Effects

Quantitative measurement frameworks. Network effects vs. virality vs. scale, multi-tenanting impact, practical KPIs (DAU/MAU by density, organic vs. paid ratios, market-by-market unit economics).

NFX Network Effects Bible

The definitive practitioner reference. Sarnoff's/Metcalfe's/Reed's Laws, critical mass, same-side vs. cross-side effects, chicken-and-egg solutions, switching costs. Continuously updated with visual diagrams.

NFX Network Effects Masterclass

3+ hour video course from operators who built 10+ companies with $10B+ in exits. 16 network effect types, case studies (Uber, Facebook, Bitcoin), Network Bonding Theory, cold start, Web3 applications.

DoorDash: Switchback Tests Under Network Effects

Why traditional A/B tests fail in three-sided marketplaces and how switchback testing with region-time randomization solves interference. Uses 30-minute time windows.

Eugene Wei: Seeing Like an Algorithm (TikTok)

Deep analysis of TikTok's success through algorithmic content discovery. Explains why TikTok's approach differs from social graph-based networks.

Sangeet Choudary: Platform Scale Blog

Blog from Platform Revolution co-author covering platform strategy, network effects, and the evolution of platform business models.

NFX Network Effects Manual: 16 Types

Most granular taxonomy: Physical, Protocol, Personal Utility, Marketplace, Platform, Asymptotic, Data, Tech Performance, Language, Belief, Bandwagon, Tribal effects. Explains why Uber/Lyft face asymptotic effects.

Rochet & Tirole: Two-Sided Markets (RAND)

Seminal paper introducing the economics of two-sided markets. Analyzes how platforms set prices when they serve distinct but interdependent customer groups.

Jean Tirole: Two-Sided Markets - A Progress Report

Nobel laureate's comprehensive survey of two-sided market theory. Covers pricing, platform competition, and welfare implications of multi-sided platforms.

Rochet & Tirole: Platform Competition in Two-Sided Markets

Foundational academic paper on platform economics. Develops theory of pricing and competition when platforms must attract multiple user groups.

Stratechery Aggregation Theory

Most cited framework for understanding internet platform dominance. Zero distribution/marginal/transaction costs, aggregator virtuous cycle, winner-take-all dynamics, platform vs. aggregator distinction.

Parker, Van Alstyne & Choudary: Pipelines, Platforms, and the New Rules of Strategy

HBR article summarizing Platform Revolution. Contrasts pipeline (linear) businesses with platform businesses and their different strategic imperatives.

Hagiu & Rothman: Network Effects Aren't Enough (HBR)

HBR challenge to the conventional wisdom about network effects. Shows why many platform businesses fail despite strong network effects.

Li Jin: Unbundling Work

How platforms are unbundling traditional employment into discrete tasks. Examines implications for workers, platforms, and the economy.

LinkedIn: The Economic Graph

LinkedIn's vision for mapping the global economy. How they use data to understand labor markets, skills, and economic opportunity.

Hagiu: Multi-Sided Platforms - From Microfoundations to Design

Academic treatment of platform design principles. Bridges economic theory with practical platform design decisions.

Hagiu & Wright: When to Open a Platform (HBR)

HBR analysis of when platforms should allow third-party developers. Framework for deciding between closed and open platform strategies.

Platform Papers (Joost Rietveld)

Academic platform research translated for practitioners. Bridges academic literature with practical platform strategy.

Andrei Hagiu: Strategic Decisions for Multisided Platforms

MIT Sloan Management Review guide to key strategic choices for platform businesses: sides to bring on board, design, and pricing decisions.

Platform Revolution Book Overview

Overview of the seminal book on platform business models. Parker, Van Alstyne, and Choudary's framework for understanding platform dynamics.

Li Jin: Building for the Creator Middle Class

Argues that the next generation of creator platforms must serve the middle class of creators, not just superstars. Framework for building sustainable creator businesses.

Bill Gurley: In Defense of the Deck

Frameworks for pitch presentations and communicating marketplace value propositions to investors and stakeholders.

Li Jin: 100 True Fans

Updates Kevin Kelly's 1000 True Fans theory for the creator economy. Argues that with higher monetization, creators can succeed with far fewer dedicated followers.

Li Jin: The Passion Economy and the Future of Work

Foundational essay on the passion economy. Explains how platforms enable individuals to monetize unique skills rather than commoditized labor.

DoorDash: Building a Successful Three-Sided Marketplace

DoorDash engineering explains the unique challenges of balancing three sides: merchants, dashers, and consumers in their delivery marketplace.

Fiverr Engineering Blog

How Fiverr structures their freelance marketplace, from gig discovery to pricing and seller success mechanics.

LinkedIn: Economic Graph to Economic Insights

LinkedIn Hiring Rate computation; partnerships with World Bank, IMF. Building infrastructure to derive economic insights from professional network data.

DoorDash: Real-Time Optimization of Delivery Operations

How DoorDash optimizes delivery operations in real-time, balancing Dasher earnings, merchant experience, and consumer wait times.

LinkedIn Engineering: Marketplace Optimization

How LinkedIn optimizes their talent marketplace to match candidates with opportunities while balancing multiple stakeholder interests.

Instacart: Building for Balance (SAGE v2)

Unique four-sided marketplace perspective (consumers, shoppers, retailers, brands). How Instacart balances all sides of their complex marketplace.

Uber Engineering

Surge pricing, marketplace design, causal inference at scale. See how researchers tackle real problems at Uber.

Bill Gurley: Going Direct

Examines how technology enables producers to bypass intermediaries. Analyzes disintermediation trends across industries from retail to entertainment.

Netflix: Keeping Netflix Reliable Using Prioritized Load Shedding

How Netflix handles overload through intelligent request prioritization and graceful degradation.

Google Research: The Tail at Scale

Seminal 2013 paper on managing latency variability in large-scale systems. Introduces hedged requests, tied requests, micro-partitioning. Won 2024 SIGOPS Hall of Fame.

Lyft: Dynamic Pricing to Sustain Marketplace Balance

Evolution of Lyft's PrimeTime surge algorithm. Explains undersupply spirals and iterative fixes for two-sided marketplace pricing.

IEEE Spectrum: The Secret of Airbnb's Pricing Algorithm

How Aerosolve handles unique inventory; neighborhood boundary mapping. External analysis of Airbnb's approach to pricing heterogeneous listings.

Airbnb: Learning Market Dynamics for Optimal Pricing

Airbnb Engineering post combining ML and structural modeling for Smart Pricing. Shows simulation-based approach to pricing.

DoorDash Engineering

marketplace analytics, delivery optimization, and experimentation. Great posts on real-time pricing and logistics.

Airbnb: Aerosolve - ML for Humans

Open-source interpretable ML showing price-demand elasticity curves. How Airbnb built interpretable pricing models that hosts can understand.

HBR IdeaCast: How to Build Dynamic Pricing That Works

Price fairness communication; cross-subsidization strategies. Harvard Business Review podcast on implementing dynamic pricing customers accept.

Airbnb: Using ML to Predict Value of Homes

How Airbnb built ML models to estimate listing value, combining property features with market dynamics and host characteristics.

Chargebee: SaaS Pricing Models Guide

Usage-based pricing, value metrics, packaging strategies — free

Lenny's Podcast: Madhavan Ramanujam

90 minutes on WTP conversations and behavioral pricing

Uber Engineering: Causal Inference at Uber

Real industry application showing how PhD-level methods translate to business problems. Covers propensity score matching at scale, RDD for dynamic pricing, and mediation modeling.

Coursera Pricing Strategy Optimization (UVA/BCG)

Price elasticity, WTP estimation, segmentation — free to audit

Monetizing Innovation (Ramanujam)

The industry bible — design products around price, not vice versa

Bill Gurley: A Rake Too Far

Classic analysis of take rates in marketplaces. Explains why high take rates invite competition and examines optimal pricing strategies for platform businesses.

The Theory and Practice of Revenue Management

Talluri & van Ryzin's comprehensive textbook. Dynamic pricing, capacity allocation, overbooking — the bible of RM.

Auctions in Ad Tech (Sanjiv Das)

GSP auctions, quality scores, AdRank — how Google/Meta ad auctions actually work. Chapter 21.

Google Research: Market Algorithms Team

Direct from engineers designing Google's auction systems. Ad exchange design, budget-constrained mechanisms, autobidding formulas, Price of Anarchy. Collaboration between Roughgarden, Tardos, and Google engineers.

Meta: Ads Fairness Variance Reduction System

Technical discussion of Total Value = Bid × Estimated Action Rate × Quality. How Meta ensures fairness in ad auctions while reducing variance.

Display Advertising with Real-Time Bidding

Free comprehensive RTB coverage on arXiv

GSP Auction Paper (Edelman et al., AER 2007)

Foundational paper on search advertising auctions

Tim Roughgarden's CS364A: Mechanism Design

The definitive free resource from a Gödel Prize winner. 20 video lectures (~75 min each) covering Vickrey auctions, Myerson's Lemma, VCG, sponsored search, combinatorial auctions, revenue-maximizing mechanisms.

Meta: Andromeda - Next-Gen Personalized Ads Retrieval

10,000x model capacity increase with sub-linear inference costs. December 2024 deep-dive on Meta's ad auction retrieval architecture.

Easley & Kleinberg: Sponsored Search Markets

Clearest pedagogical treatment of online ad auctions. VCG from 'harm principle,' GSP mechanics, GSP vs VCG comparison with worked examples, why truth-telling isn't dominant in GSP. Perfect for understanding Google/Facebook ads.

Kevin Leyton-Brown's VCG Mechanism Lectures

Structured theorem-proof format with worked examples. VCG formal definition, DSIC proofs, Clarke pivot rule, budget balance, shortest path auctions. Shows exactly how second-price sealed-bid is VCG special case.

Dirk Bergemann's Yale Courses

Yale courses on information economics, mechanism design, and dynamic auctions from leading auction theory researcher

VisuAlgo

Animated algorithm visualizations — sorting, graphs, DP

TheAlgorithms/Python

200+ algorithm implementations in Python — reference code

LeetCode Explore: Data Structures

Structured practice cards with solutions

Problem Solving with Algorithms & Data Structures (Python)

Free interactive textbook — visualizations and runnable code

Real Python: Data Structures

Practical guide with Python-specific implementations

USF Data Structure Visualizations

Interactive animations — see how trees, heaps, and graphs work

freeCodeCamp: Algorithms Course

8-hour free video course — clear explanations

Abdul Bari (YouTube)

Exceptional whiteboard DSA explanations

DataLemur

Real DS interview questions with business context

NeetCode

Curated LeetCode roadmap organized by pattern. Video explanations that actually make sense. The modern way to prep for coding interviews.

The Missing Semester (MIT)

Command line, Git, debugging, shell scripting. The CS skills they don't teach in econ PhD programs but you absolutely need.

Playwright for Python

Modern browser automation (faster than Selenium)

Coding for practitioners

Built specifically for econ researchers

Python for Econometrics

Kevin Sheppard's comprehensive intro for researchers. NumPy, pandas, statsmodels, and econometric applications.

Automate the Boring Stuff with Python

The best free Python book for non-programmers. Web scraping, Excel automation, file management — practical skills for data work.

Python Data Science Handbook

Free reference for NumPy, Pandas, Matplotlib

QuantEcon: Discrete State Dynamic Programming

Gold standard for DP in economics. Bellman equation, value/policy iteration, contraction mapping proofs, stochastic optimal growth. Runnable Jupyter notebooks implement DiscreteDP class.

Coding for Economists

Practical guide by A. Turrell on using Python for modern econometric research, data analysis, and workflows.

Matteo Courthoud's BLP Demand Estimation

Exceptionally clear BLP from first principles. Share inversion, nested fixed-point step-by-step, instrument selection (BLP, Hausman, cost shifters), GMM estimation. Python implementation included.

SQLBolt

Learn SQL with interactive exercises. No setup required — run queries right in the browser. Perfect for beginners.

Mode SQL Tutorial

Interactive SQL lessons from basic to advanced. Great for learning JOINs, window functions, and subqueries with a real database.

LeetCode SQL 50

50 essential SQL problems to master for interviews. CTEs, window functions, and common patterns used at FAANG.

DuckDB Documentation

Modern in-process SQL database. Runs on your laptop, reads Parquet directly, and is perfect for analytics. The new pandas killer.

SELECT Star SQL

Interactive book teaching SQL through meaningful analysis

8 Week SQL Challenge (Danny Ma)

8 business case studies with CTEs and window functions

PyMC-Marketing Documentation

BG/NBD and Gamma-Gamma CLV tutorials

Statistical Rethinking

Richard McElreath's Bayesian approach to statistics. PyMC3 translations available. The book that changed how many think about inference.

PyMC Labs Blog

Bayesian causal inference done right. MCMC, probabilistic programming, and causal models from the PyMC team.

Seeing Theory (Brown)

Beautiful interactive visualizations for building intuition

Scipy.stats Documentation

Reference for distributions and tests

Evan Miller: Formulas for Bayesian A/B Testing

Mathematical foundations for Bayesian approaches to A/B testing. Derivations of exact formulas for posterior probabilities and expected loss.

StatQuest with Josh Starmer

Visual explanations of cross-validation, regularization, gradient boosting, PCA, and bias-variance tradeoff. 675,000+ subscribers. Fills conceptual gaps that course-based learning misses.

Money Stuff (Matt Levine)

Daily newsletter making complex financial mechanics accessible. 300,000+ subscribers. Wall Street, M&A, tech IPOs, and securities law explained with wit.

Noahpinion (Noah Smith)

applied analytics, AI, innovation, growth. Deep dives with data, accessible to non-specialists. The researcher's tech newsletter.

GenAI for Econ Substack

Anton Korinek's Substack newsletter with updates on LLM capabilities for economists and practical applications in research.

Koen Pauwels: Marketing and Metrics

Editor-in-Chief IJRM, VP of Practice at INFORMS, consultant to Amazon/Microsoft/Unilever. Bridges academic marketing science and industry practice with weekly LinkedIn newsletter.

MKT1: B2B Marketing Frameworks

Emily Kramer (former VP Marketing Asana/Carta) serving 45,000+ subscribers. Lenny Rachitsky calls it his '#1 favorite marketing newsletter.' Krameworks templates for marketing measurement.

Adam Fishman Newsletter

Ex-Patreon/Reforge. Growth loops, product strategy, and how to structure product analytics.

Marketing BS: Strategic Marketing Newsletter

#1 marketing newsletter on Substack (21,000+ subscribers, 40%+ open rates). Edward Nevraumont (former VP Expedia, CMO General Assembly) challenges conventional marketing wisdom.

Bill Gurley: How to Miss By a Mile (Uber TAM)

Analysis of TAM (Total Addressable Market) estimation errors. Explains why most TAM analyses are flawed and how to think about market sizing for tech companies.

Kuang Xu Newsletter

Stanford GSB Professor bridging OR research with AI strategy and experimental design. Posts like 'The Importance of Being Modest' combine academic rigor with business applications.

SemiAnalysis (Dylan Patel)

Deep technical analysis of semiconductor economics and AI hardware. 200,000+ subscribers. Ben Thompson's 'most important and most-cited resource'.

Matt Levine: The Crypto Story (Businessweek)

40,000-word feature explaining the entire crypto ecosystem. Exemplifies Levine's ability to explain intricate systems accessibly.

The Diff (Byrne Hobart)

Daily (5x/week) analysis of inflection points in finance and tech. 47,000+ subscribers including '1.5% of the Forbes 400'. Matt Levine for mental model geeks.

Apricitas Economics (Joseph Politano)

Data-driven macroeconomic analysis with exceptional visualization. Noah Smith calls it 'one of the best econ data blogs'. Labor markets, inflation, industry economics.

Not Boring (Packy McCormick)

#1 Business newsletter on Substack with 239,000+ subscribers. Long-form startup deep-dives often exceeding 10,000 words. 'Ben Thompson meets Bill Simmons'.

Economic Forces (Albrecht & Hendrickson)

Chicago-style price theory for modern audiences. 23,000+ subscribers. 'By far the best newsletter on economics' per Anton Howes.

How Superhuman Built an Engine to Find PMF (First Round)

Operationalizes Sean Ellis's '40% very disappointed' survey into a systematic process. How Superhuman went from 22% to 58% PMF score using segmentation. Most referenced First Round article.

Slack's 2000 Messages Activation Metric

Documents Slack's activation discovery — after 2,000 messages sent per team, 93% remain active. How they identified this leading indicator. Conversion rate significantly above 5% SaaS average.

Adam Fishman Newsletter

Ex-Patreon/Reforge. Growth loops, product strategy, and how to structure product analytics.

Atlassian: How to Write Product Requirements

Practical guide to writing PRDs in an agile environment from the makers of Jira and Confluence. Includes templates and explains modern, lightweight documentation.

Guide to Product Metrics

26 metrics across AARRR framework: activation, retention, LTV, NRR, Quick Ratio, PMF Score explained

Sequoia: Data-Informed Product Building

Metric hierarchies, North Star metrics, and building data-informed products. The definitive framework for product metrics.

Measuring Product Health (Sequoia)

Definitive guide to growth, retention, stickiness & engagement metrics: DAU/MAU, Lness, cohort curves, Quick Ratio

A Quantitative Approach to Product-Market Fit (Tribe Capital)

The foundational text on growth accounting. MAU growth accounting AND revenue growth accounting. Quick Ratio, Gross Retention, Net Churn explained by the team that pioneered it.

Ultimate Guide: Activation (Aakash Gupta)

Traces activation history from Facebook's 2008 growth team, including Chamath's '7 friends in 10 days' discovery. The Setup → Aha → Habit framework with data-backed examples.

Teresa Torres: Opportunity Solution Trees

Product discovery coach who has trained 17,000+ PMs. The Opportunity Solution Tree framework connects business outcomes → customer opportunities → solutions → experiments.

Great Learning Insurance Analytics Course

Free course covering insurance analytics fundamentals including customer segmentation, claims prediction, and fraud detection with Python implementations.

IEEE-CIS Fraud: 1st Place Solution (Chris Deotte)

Kaggle Grandmaster, 262 features, RAPIDS GPU

scikit-learn: Outlier Detection

Isolation Forest, LOF, One-Class SVM comparison

Stripe Engineering

Payment economics, fraud detection ML, financial data infrastructure. Building economic infrastructure for the internet.

Fraud Detection Handbook (ULB)

From the team that created the Kaggle dataset — rigorous methodology

PayPal: Graph Database for Fraud

Real-time fraud ring detection

Stripe: ML for Fraud Protection

The definitive intro: features, precision-recall tradeoffs, break-even calculations

LinkedIn: Defending Against Abuse at Scale

4M+ TPS, multi-layer defense architecture

Netflix: RAD Outlier Detection

Robust PCA at terabyte scale

Google Research: Self-Supervised Anomaly Detection

Contrastive learning, CutPaste algorithm