Selected work focused on customer behaviour, experimentation, revenue, business intelligence and reliable analytical decision-making.
Evaluated a 64,000-customer randomised email experiment to identify the strongest treatment, estimate incremental spend and contribution profit, test segment heterogeneity and determine whether personalised targeting improved on the best fixed policy.
Designed an end-to-end analytical workflow for an e-commerce retailer using PostgreSQL, Python and Power BI. The project combines customer segmentation, churn-risk scoring, value prioritisation and cohort retention analysis.
A temporal churn case study designed around cohort behaviour, leakage-free feature engineering, calibration, intervention capacity and retention economics—not only model accuracy.
An event-level GA4 and BigQuery case study investigating user journeys, conversion leakage, channel performance, device behaviour and the experiment opportunities most likely to improve revenue.
Earlier projects document my technical development across dashboards, exploratory analysis and predictive modelling. They remain available on GitHub, but the portfolio now prioritises deeper business case studies.