Klaviyo uses machine learning to automatically generate per-customer predictions — including predicted customer lifetime value (CLV), churn risk score, expected next order date, and predicted spending — for every qualifying customer profile on the platform. This capability has been part of the platform since at least 2017 and is embedded in all paid marketing plans.

Klaviyo's predictive analytics system ingests each customer's full purchase history, engagement signals, and cross-brand behavioral data. It applies machine learning and Bayesian inference techniques to output four core scores per profile: predicted CLV (next 365 days of spending), historic CLV, churn risk probability (0–100%), and expected date of next order. Models are retrained at least weekly. The system requires a minimum of 500 customers with order history to activate. Klaviyo confirms these predictive features are core to the platform and cannot be opted out of. Advanced CLV customization (adjustable prediction windows) is available on higher-tier Marketing Analytics and Advanced KDP plans.