I would not put a lot of time into a computer algorithm generated value.  

That's one of the major problems with Zillow algorithms....they have never been inside the house so their Zesstimates are never accurate.

The CEO was heard to say "We are losing $1,143 on every deal but we'll make it up in volume" ;-)"Zillow Shares Fall After Forecasting Home-Flipping Losses"Zillow Group Inc. shares fell as much as 14% in after-market trading after the company forecast that its new home-flipping business would lose as much as$80 million in the third quarter before interest, taxes, depreciation and amortization.The company generated $249 million in second-quarter revenue from Zillow Offers, an algorithm-driven spin on home-flipping, and received requests from 70,000 consumers who wanted to sell their homes to Zillow over the internet.

I do think bad data is better than no data, with regards to the Zestimate...My 'give them the benefit of the doubt' in their algorithm is that perhaps us folks that are interested enough to actually click on the 'Edit Facts' button (us) do correlate to being relevant to adjusting value - vs those that don't. 

Still I question if the algorithms can get the price right.

Conwell zestimate is an algorithm that evaluates off certain criteria-a human set of eyes has not been set on the property...

A Zestimate is an output from a machine learning algorithm.

We all know there can be wild error bars on that price however there is some value in an algorithm pooling data from home sales.   

I'm not sure, check with a moderator, or it could be an automatic process based on a computer algorithm.

Zillow doesn't entirely have their algorithm sorted out on them.If the house is older but could be livable for a minimum reasonable amount, you should fix it and put a renter in it.If it's old and too rough, you should tear it off and replace it with a brand new manufactured home and sell that.