Got Some Questions While Building Sales Forecast Models...

Hey I'm really just a newbie in ER so please bear with me.

My role is also different from traditional ER associate. We have a data science team that supports research (I guess many banks do) and I have been trying to build some sales forecast models for our analysts. Initially the analyst was just curious about how specific factors like weather would impact the sales.

I'm from stats & data science background so I thought we shouldn't just run a simple correlation before we filter out impact or contributions from other factors. Also I think the relationship could be non-linear so I kind of adopt some approach that is normal in data science world but maybe too complicated to interpret for our analysts. But they haven't show any objection so far and even give me some guidance in terms of incorporating more features into the model.

Sometimes I ask myself would this be too much? I know that no model would be correct and I personally don't have very strong confidence in the model for various reasons (limited sample size, sensitivity to the input data, etc.). I heard that in ER people actually don't take the price target that seriously. Then does the accuracy of my model or "Forecasting" this whole thing really matter for people in ER?

I also learned in CFA Level 1 that an analyst should have thorough understanding about his/her model. Then am I supposed to teach them the AI stuff which might be too difficult to understand for them? Or maybe they won't take my model seriously at all? I wonder what ER people usually do and think when they are provided with this sort of "predictions" from say 3rd party or data vendor or their colleagues?

Sorry for being verbose. Really appreciate your time and help. Many thanks in advance.