Clarity over jargon. The same standard on stage as in the boardroom.
I speak on data science, AI strategy, and decision-making. The goal is frameworks audiences can actually use, not a tour of the hype cycle.
My talks are designed for executive audiences navigating real decisions, not theoretical frameworks. The focus is on clarity, tradeoffs, and actionable thinking that survives contact with a boardroom.
The talks I've delivered most often, and the ones that have generated the most conversation afterward:
Executives need a diagnostic framework to separate genuine AI innovation from expensive vendor theater. This talk categorizes AI into five functional types (Rules-Based, Statistical, Machine Learning, Deep Learning, and Generative) and gives leaders a practical tool for evaluating claims and allocating budgets. Audiences leave knowing what questions to ask rather than what buzzwords to repeat.
For ExecutivesMost corporate testing programs fail in predictable ways: testing the wrong things, using irrelevant metrics, or rerunning the same experiment until it produces the hoped-for result. This talk is a strategic audit of experimentation culture, designed to help organizations distinguish rigorous testing from confirmation theater.
For Executives For PractitionersIn a subscription economy, "hours streamed" is an insufficient content metric. This talk presents a framework for estimating the true incremental value of individual content assets — separating titles that drive acquisition and retention from titles that merely get watched because the audience is already there. Uses counterfactual modeling and Shapley value approaches.
For Executives For PractitionersA subscriber analytics talk with a deceptively simple premise: why do acquisition and retention teams always seem to be working against each other? The talk unpacks the mathematical relationship between growth and churn, explains why a single churn rate number is almost always misleading, and introduces the Integrated Churn Rate (ICR), a CAGR-style metric that captures the full shape of a retention curve in one number executives can actually use.
For Executives For PractitionersMultiple appearances across several years, covering the transition from legacy ratings to subscriber-centric lifetime value models. Discussions focused on synthesizing disparate data from streaming platform partners (Apple, Amazon, Roku) into a coherent global subscriber strategy. Also served as a track chair, shaping the analytical agenda for the conference's data science programming.
Presented on the philosophy of the control group: the argument that sophisticated modeling is secondary to a rigorous counterfactual for measuring true incremental impact. A case against complexity for its own sake, and for the discipline of asking "compared to what?" before building anything.
Regular contributor and speaker for the All Things Insights platform, covering AI strategy, decision science, and the practical realities of running analytics organizations in large enterprises. The column is available at allthingsinsights.com.
Two appearances: a panel on AI and consciousness (2025), and a 2024 co-presentation with a visual artist on the parallels between good art and good decision science, two disciplines that both require knowing when to commit, when to iterate, and when to trust pattern recognition over explicit rules.
Presented live for SSP International alumni, then made available publicly through SSP's YouTube channel. Watch on YouTube →
Available for keynotes, panels, workshops, and executive briefings on data science leadership, AI strategy, decision-making, marketing analytics, and data culture.
A speaker one-pager with topic descriptions, format options, and technical requirements is available on request. Email contact@michaelbagalman.com for availability and details.