Enterprise AI
Propensity model
AI estimating how likely someone is to do something, such as buy a product.
Example
An AI system estimates whether a customer is likely to accept an offer.
Why people use it
It helps estimate who is likely to take a particular action.
What you'll hear
“Who seems likely to accept the offer?”
What this means for you
Test proposed actions independently rather than treating likelihood as causation.
Can you control it?
Sometimes
Sometimes. Your choices depend on the tool and your access. The settings available to an everyday user may differ from those available to the people running it.
Common questions
- Is a propensity score proof that an intervention will work?
- No. Predicting behavior differs from estimating the effect of an intervention.
- Can a high score reflect someone who would act anyway?
- Yes. Likelihood of acting does not show that a message or offer caused the action.
- Can the score become stale?
- Yes. A person's needs or behavior can change after the estimate is made.