One of the big topics of discussion recently is using the results of lift tests to calibrate MMM.
This is something we champion at Recast, and regularly advise our clients to validate our model by running lift tests to build internal confidence in the model. We welcome the transparency.
You can’t test everything, and MMM is that glue that connects everything together in between tests. When the model is correct we can use it to extrapolate the incrementality of things you haven’t tested recently.