William Gieng
Staff Data Scientist, Qualtrics
Portfolio
Staff-level data scientist focused on causal inference, experimentation, and the translation of statistical evidence into senior leadership decisions.
Case studies translating ambiguous business questions into defensible analytical decisions.
01 — Causal inference
A $300M renewal book rested on an unproven thesis. Part I establishes a +12.9pp causal lift with five converging estimators. Part II maps where the effect lives and where spending stops.
Decision impact
Pooled lift adopted as the standing planning value for renewal forecasting; segment results feed the executive churn review and a board-level action plan.
02 — Retention
When a competitor disruption drove a 280K subscriber spike, was it real growth or borrowed users? Three independent lenses on the same question.
Decision impact
Reframed leadership conversation from acquisition volume to activation quality and retained LTV.
03 — Prediction
Can first-month behavior separate "binge and leave" users from habit builders? Acquisition channel, billing, and engagement decomposition.
Decision impact
Surfaced channel-level churn risk and "binge and leave" patterns informing onboarding and acquisition spend.
Long-form essays on causal identification, measurement, and the discipline behind decision-grade analysis.
Separating overlapping churn drivers when promo expiry and use-case completion arrive together.
Connecting LLM summarization failures to causal identification discipline.
Why LLM-extracted variables carry a selection, timing, and measurement footprint the downstream model never sees.
Public notebooks demonstrating the methods behind the case studies and writing.
DiD, synthetic control, RDD, and ITS on a synthetic subscription pricing dataset, with LTV translation and breakeven price calculation.
Three-stage LLM pipeline (extract, synthesize, audit) for decision-grade meeting analysis with explicit grounding and bounded fabrication controls.
Pipeline that extracts and categorizes churn reasons from unstructured text using weak supervision, transformer fine-tuning, and prompt-engineered LLM summaries.
A LinkedIn series on the framing decisions that determine whether analysis is causal.
A single churn rate cannot tell you which lever to pull. The aggregate decomposes into distinct mechanisms, from fit failure to competitive loss, each pointing to a different decision. The number is the start of the analysis, not the end of it.
Activation and retention are related but do not always move together. A lift can be real improvement, a pull-forward effect, a selection effect, or short-term behavior that fades. The causal question is whether the change kept users who otherwise would have churned.
Teams choose metrics because they are available, stable, and easy to move, then treat the proxy as the outcome. Pushing on a proxy is a treatment effect problem. The question is whether the effect transfers to the outcome it was supposed to represent.