Specifying non-default priors

[mthomas-ketchbrook]

Hi,

Similar to many others I'm sure, our team (myself & @patrick-immensityforartists) is fitting a DelayedSaturatedMMM model while following the pymc-marketing documentation.

We are struggling with the interpretation around specifying priors ourselves (as discussed briefly in this section), due to both the min-max scaling that takes place as well as the adstock/saturation transformations.

If we were looking for our prior to communicate that we expect on average $5 in sales for every $1 spent on channel A, what transformations (if any) would need to take place in the 'mu' parameter for our prior specification for channel A (e.g., using a lognormal distribution specification)? Do we need to apply min-max scaling to the mu and sigma values of our priors?

I appreciate any information that can help point us in the right direction.

Thank You!

[cetagostini]

Hello @mthomas-ketchbrook

Take a look here: #617 (reply in thread)

This should help you to understand the function and estimate the prior, let me know if you have any other questions!

Happy to help 👍