What Are Propensity Scores in Marketing?
Propensity scores in marketing are numerical estimates of how likely a person, account, or audience segment is to complete a specific action. That action may be making a purchase, requesting a demo, renewing a subscription, becoming a qualified lead, or responding to a campaign.
The score is usually calculated from patterns in behavioral, demographic, firmographic, transactional, or engagement data. A higher score indicates a stronger predicted likelihood of the selected outcome, but it does not guarantee that the action will happen.
What Do Propensity Scores Measure?
A propensity score measures the probability of one clearly defined outcome. Teams may create separate scores for purchase likelihood, churn risk, lead qualification, product adoption, or subscription renewal.
Common inputs include website visits, pricing-page activity, repeat sessions, email engagement, advertising interactions, CRM stages, purchase history, and product usage. The data used should be relevant to the action being predicted.
| Input type | Example |
| Behavioral | Product pages viewed or repeat visits |
| Engagement | Email clicks or campaign responses |
| Transactional | Previous purchases or average order value |
| CRM | Lead stage, account status, or sales activity |
| Firmographic | Company size, industry, or role |
The quality of the score depends on accurate data and clear outcome definitions. Missing events, inconsistent CRM records, or low-quality conversion data can make a score appear precise while producing unreliable recommendations.
How Are Propensity Scores Used in Marketing?
Marketing teams use propensity scores to prioritize audiences, adjust messaging, trigger lifecycle campaigns, and support sales follow-up. Instead of treating every user or lead equally, teams can focus resources on people who show stronger predicted conversion potential.
For example, a paid media team may create remarketing segments based on conversion propensity. A sales team may prioritize high-scoring leads, while a lifecycle team may send different messages to users with low, medium, or high purchase likelihood.
Propensity scores also support broader predictive audience targeting by helping marketers decide which audiences should receive more budget, personalized offers, or additional follow-up.
Scores should be tested against actual outcomes and reviewed regularly. Customer behavior, campaign conditions, and market demand can change, causing older scoring rules or models to become less reliable.