Predictive Analytics Overview #
DMSI Analytics includes built-in predictive models that help you look forward, not just backward. Instead of only seeing what has happened, predictive analytics tells you what is likely to happen.
Accessing Predictive Analytics #
Go to DDLS > Predictive Analytics in your WordPress admin.
Three Predictive Models #
| Model | Purpose |
|---|---|
| Churn Predictor | Scores each customer 0-100 for churn risk |
| LTV Predictor | Forecasts per-customer lifetime value |
| Revenue Forecaster | Projects revenue 1, 3, 6, and 12 months ahead |
When Models Run #
All three models run automatically as part of the daily metric cron at 2:00 AM. You do not need to trigger them manually.
For on-demand recalculation, the Predictive Analytics page includes a Retrain Models button. Use this after making significant changes to your product pricing or business model.
Model Accuracy #
Predictive models improve over time as more historical data accumulates. Models require at least 30 days of data to produce meaningful results. With less than 30 days of data, you will see a notice indicating low confidence.