Move beyond point estimates and gut feelings. Our Bayesian framework gives you honest uncertainty ranges so you can make decisions with confidence.
When you run many models and only report the "good" ones, you're painting targets around arrows.
Traditional approaches often involve running dozens of model variations and selecting the one with results that "make sense." This feels rigorous but actually destroys statistical validity.
Watch the dartboard: each throw represents a model specification. Only the bullseyes (models with "good" results) get reported. The misses? Quietly discarded.
The result: Your reported 100% accuracy is an illusion. The real accuracy is much lower, but you'll never know because you only see the "winners."
The darts that miss are models with "unrealistic" results—quietly discarded.
🎯 Real accuracy: 15% | Reported: 100%
Choose the right level of complexity for your measurement challenge.
The foundation. Measures how media drives a single outcome (sales, leads, etc.) with proper uncertainty quantification.
Captures how media works through intermediate steps like awareness or consideration before driving sales.
Measures interactions between products—how promoting one SKU affects others (cannibalization or halo effects).
Traditional variable selection (stepwise regression, p-value hunting) is a form of specification shopping that invalidates inference. Our framework provides Bayesian alternatives that quantify uncertainty about which variables matter.
But variable selection is not a general-purpose tool. It should only be applied to precision control variables—never to confounders, mediators, or your media variables themselves.
Posterior inclusion probabilities quantify variable importance
Watch how prior beliefs combine with data to produce honest uncertainty ranges.
Promoting one product affects others. Ignoring this leads to inflated ROI estimates.
Everything you need to go from your first model fit to a quarterly closed-loop measurement program.
Install the framework, fit your first Bayesian MMM, and walk through a complete code example.
Step-by-step guidance for building statistically sound MMMs - priors, hierarchy, diagnostics, and iteration.
How to read MMM outputs, communicate uncertainty, and translate posteriors into confident budget decisions.
Wire experiments and the MMM into one self-correcting cycle - EIG, EVOI, and the virtuous loop.
The disciplined Bayesian process: priors, prior predictive checks, fitting, posterior diagnostics, and PPCs.
DAGs, do-calculus, and counterfactuals - the causal scaffolding behind every defensible MMM specification.
When to include precision controls, why confounders are non-negotiable, and how shopping for variables backfires.
The risks of specification shopping and what defensible measurement looks like for agencies and brands.
Interactive workflow demonstrations, scenario analyses, and an example MMM report from a Q4 2025 fit.
Math and implementation details for every model in the framework - standard, nested, multivariate, and combined.
Module-by-module API docs (Sphinx). Browse the source on GitHub or build locally with make html in docs/api.
Definitions of every term used across the docs - adstock, EIG, EVOI, MFF, ITT, MDE, ROPE, and more.
Common questions about MMMs, Bayesian methods, and the trade-offs the framework makes deliberately.
Current version, API stability tier per module, versioning policy, and release notes.
Background on the framework, its design principles, and the audiences it serves.
The framework is open source and ready to use. Start with our documentation or dive straight into the code.