Churn Prediction Model — B2B SaaS Platform
18%→9%
Annual revenue churn rate
90d→14d
Average churn signal lead time
67%
High-risk accounts retained after CS intervention
£3.1M
ARR retained in first year
The Challenge
A B2B SaaS company with 4,200 accounts was losing 18% of ARR annually to preventable churn. Customer success managers had no early warning system — churn was typically identified at renewal, 90 days too late for meaningful intervention. High-value accounts were departing with no prior signal.
What We Built
We built a churn prediction model trained on 3 years of product usage, support, billing, and engagement data. The model scores every account weekly on 90-day churn probability, segments accounts into risk tiers, triggers CS workflows for high-risk accounts, and surfaces the specific feature gaps and engagement signals driving each risk score.
“We went from finding out accounts had churned to having 2 weeks to do something about it. The model paid for itself in the first quarter.”
— Chief Customer Officer
Want a Result Like This?
SmartPath AI builds and deploys production AI systems for enterprises. Schedule a strategy session to discuss your specific use case.
Schedule Strategy Session