Predictive scoring

What is predictive scoring?

A predictive model learns from past customer data: purchases, interactions, engagement over time. It then calculates a score for each contact, for example a churn probability, an expected order frequency or an expected future revenue, a key input for customer lifetime value. Unlike rules set by hand, the scores update as behaviour changes, and they can drive your activation scenarios.

Predictive scoring lets you anticipate instead of react. You can protect customers at risk before they leave, invest more in those with high potential and avoid over-messaging contacts who are unlikely to respond, which also protects deliverability.

Actito brings machine-learning scores such as churn probability, anticipated order frequency and expected future revenue directly into your data model through scores tables. You use them like any other data: to build segments, trigger scenarios or personalise offers. An engagement score, calculated automatically for every contact, complements them. Learn more about AI in Actito.