Zillow Adjustment Factor

[Erik Sherburne]

I'm looking at a SFH portolio. The selling price of the portfolio is discounted based off the Zillow price. 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've toured all the properties, took pics, made notes, the usual stuff and as I tried to come up with my own value I came up with a scale to consider improvements needed vs the current Zillow price.

Scale	Adjustment Factor
1	Tear Down	-60%
2	Major Repair	-35%
3	Significant updates needed	-15%
4	Cosmetic updates needed	-5%
5	Zillow value - intentionally above average as Zillow factors in "sale ready" condition	0%
6	Major improvements	10%
7	High end value added everything	25%

You'll notice that I consider "Zillow value" to be a 5 (greater than average) as my assumption is that most properties that are for sale are also above average outliers (on either side) factor out in to the "noise".

Thoughts?

[Mike McCarthy]

if you’re looking at row or tract houses where they are similar, Zillow can do pretty well. But outside that, it’s hard to trust the Zestimate. They just don’t have enough data on the finishes of the estimated house nor the houses that feed into that estimate. All they see in many cases is 3br 2ba and xxx sq ft (which is usually wrong anyway).

Just in my rather limited use, I’ve seen it literally $100K+ off from other comps.