Hello again! I think I’ve realized something I was doing wrong 
I think I was trying to fit a single coefficient to several time series with different behavior for my priors, thus quite logically not finding a workable result. What I meant to do was fit the coefficient to the aggregated level time series, and use that as the prior for the next hierarchical level!
Thus I was wondering how to do this best. Can I run a separate model at the most aggregated level, and use the resulting coefficients as priors? Or aggregate the data in the model itself? Or make separate data sets for the aggregated series and include it in the same model as ‘Observed’? I was looking at the Potential function but I can’t tell if it would be a good use for it.
Thanks again as always!