at my company, we have a product level forecast that we run through a model which pulls out an hours number for our retail outlets.
We do this by binning various products into categories and give them a time value lets call this tmv (timed minute values). So product coffee gets a tmv of 0:50 s per quantity.
we then do a simple product * tmv and aggregate up while accounting for certain constraints, such as minimum hours required to run a shop floor, management hours and so forth.
when I inherited this beast, it was all written in excel which I ported into SQL and Python, although a lot of work has gone into it, there is little statistical modeling, optimization, or basic ML applied. Although, I'm not sure what could be done to better optimize this.
I wonder what more to-date tech companies do to allocate their work-force in a retail environment.