Thread
Causal ML
Estimating who would change their behaviour, and turning that into a decision about who to contact.
Predictive models rank people by what they are likely to do. Causal ML estimates what would be different if you acted, which is a harder question and usually the one that matters. It spans two disciplines on purpose: the estimation belongs to incrementality, the policy it feeds belongs to decisioning, and a model validated offline against historical outcomes has demonstrated neither.
Decisioning
- Planned
Prediction Is Not Persuasion
A high-propensity customer may convert without you. A high-risk one may be unreachable. The best target is whoever would change their mind.
- Planned
The Case for Doing Nothing
Contact fatigue, capacity, and cost mean no-action is a real treatment arm. Why it belongs in every decisioning system by default.
- Planned
Uplift Modeling in Practice
Estimating heterogeneous treatment effects when you have a randomized holdout, and how to validate a policy prospectively rather than offline.
- Planned
Why Agentic Marketing Still Needs Holdouts
Automating a decision does not exempt it from measurement. Logged decisions, persistent controls, and causal evaluation for autonomous systems.
- Worked example
Broad Nurture, Propensity, or Uplift?
One trial-to-paid nurture sequence, four targeting policies, and a randomized holdout. Compared on incremental conversions rather than model accuracy, which is the only comparison that changes the answer.
Systems
- System Planned
Next-Best-Action Policy
Turning scores into decisions under real constraints: capacity, contact fatigue, margin, cooldowns, and the explicit option of doing nothing.
- System Planned
Governance for Agentic Campaigns
Logged decisions, persistent controls, consent enforcement, and causal evaluation when an autonomous system is choosing the action.