Causal Inference for Real World Decision Making

Workshop agenda

• Causal Thinking for Real-World Decisions

  Learn why causal inference matters when A/B tests are unavailable and how counterfactual thinking helps answer “what caused what?”

• Turning Business Questions into Causal Estimands

  Convert vague stakeholder asks into clear treatments, outcomes, populations, time windows, and estimands like ATE, ATT, and CATE.

• Regression Adjustment for Product Feature Impact

  Estimate feature impact by comparing naive and adjusted results while accounting for observable confounders in business data

• Lunch break

  Pause, recharge, and return ready for hands-on causal model implementation

• Matching Methods for Marketing Campaign Lift

  Build comparable treatment and control groups to estimate incremental impact from targeted marketing campaigns

• Difference-in-Differences for Product and Policy Rollouts

  Measure product, policy, or regional rollout impact using pre/post trends and treatment/control comparisons

• Short break

  Short reset before the second half of hands-on modeling and case work

• Propensity Scores and IPW for Churn Interventions

  Use propensity scores and inverse probability weighting to estimate retention or churn intervention impact

• Heterogeneous Treatment Effects and Uplift Thinking

  Move beyond average impact to identify which users, customers, or segments benefit most from a treatment

• Capstone: Method Selection and Stakeholder Communication

  Choose the right causal method for business cases, then explain assumptions, uncertainty, and recommendations clearly to stakeholders