They will give you an overall “tenant score” which might make your choices easier as they run everything thru their algorithms and so far they have been right with the folks they recommend for me. 

I'm curious if underwriters have any algorithms or programs that take care of this automatically based on a few inputs or if they largely pull comps manually? 

I wouldn't rely too heavily on what Zillow says since it's never seen the house and it just based on algorithms.

No algorithm or computer will ever be able to out-evaluate personal knowledge on a property. 

Researching past sales, price per sq foot, and theoretical bank algorithms can be a fun past time, but none of that stuff matters.

Please upvote my post for the algorithm :-)You can PM me if you have questions about the methodology (it should be updated as 2019 census data is in).

Those numbers are done through a generic algorithm.  

I will pre-approve to help algorithms but I will reply to their question with the answer that they obviously do not want to hear.

A chain reaction of selling as the trading algorithms feed into each other (why markets fall faster than they rise). 

It's not really a good source of lead generation because if you are identifying it by sight as a property of value, it's likely already been identified months, if not years before by a predictive analytics algorithm.