Modelling guides
Work from the question and data towards a decision. Each guide states the supported API and the assumptions needed to interpret its outputs.
| Step | Guides |
|---|---|
| Define the question | Workflow, identification |
| Prepare observations | Data and lifecycle, scaling and pooling |
| Specify the model | Priors and transforms, fixed effects, correlated random effects |
| Add structure or evidence | Holidays, calibration, custom effects |
| Evaluate the fit | Inference, diagnostic gates, blocked holdout, prior sensitivity |
| Interpret outputs | Contributions and ROAS, tables and plots |
| Compare actions | Scenario recipes, optimisation |
| Operate the runner | YAML configuration, output files, AI advisor |
| Explore separate capabilities | Predicted incrementality (PIE), optional integrations |
| Diagnose a failure | Troubleshooting |
| Resolve common questions | FAQ |
See the API reference for public entry points and the methodological references for supporting literature.