Short answer
A projected churn increase means the forward-looking model expects customer losses to worsen in the next period. The first question is whether the forecast reflects real risk or just noisy inputs and temporary conditions.
What it usually means
At its best, the signal gives the team time to act before retention damage lands in reported numbers. At its worst, it is driven by unstable seasonality, small-sample cohorts, or payment problems being misread as broad product churn.
Main causes
- Weak cohorts or at-risk segments are entering their typical churn window.
- Payment failures, dunning weakness, or billing friction are raising predicted losses.
- Renewal timing and seasonality make the next period look worse than the long-term pattern.
- Forecast inputs are too noisy or too concentrated to support a strong conclusion.
What to check next
- Compare the projection with Churn Spike Detected and Dunning Recovery Rate Is Dropping.
- Validate historical quality with Customer Churn Rate Formula and NRR.
- Inspect the forward view in Revenue Forecasting Demo.
Product angle
Predictive churn alerts only create value when they explain which drivers lifted the forecast. Otherwise the team sees a scary number without knowing whether to fix product, payments, or the model itself.