We just put Zillow‘s value Zestimates™ to the test in two Sandburg Village buildings — 1255 and 1355 N Sandburg — and it appears to earn an A to an A-plus rating.
Now Sandburg’s a bit of a lay-down case for estimating values accurately: hundreds of commodity units that vary little from one to another.
Only a few units in these buildings are currently listed with brokers, so our sample was small, consisting of only 9 units in the MLS. Six of those were listed with “brand-name” Realtors, and 3 with “no-name” Realtors (you know who you are).
If you exclude the three units listed with no-name Realtors from the results, Zillow’s value estimates are, on average, ½ of 1% lower than the average MLS asking price for the units. The variance in individual value estimates was tight, ranging from 2.76% below to 2.03% above the asking prices.
Add the three excluded units back to the results and Zillow estimates were, on average, 1.15% below the MLS asking price. The variance was more striking: Zillow’s value estimates were between 9.84% lower and 8.36% higher than the asking prices set by the Realtors.
It’s difficult to know exactly what conclusions to draw from these numbers, but I’ll venture a few.
First, a homeowner who sets an asking price in these buildings based on the Zestimate (without even using Zillow’s tools to refine estimates) is likely to set the same price that a brand-name Realtor would set.
Second, a homeowner basing pricing purely on a Zillow esimate is likely to price, on average, more accurately than a no-name Realtor.
Third, if Zillow’s value estimates are accurate — and that’s a decent-sized if — even the brand-name Realtors are probably setting the asking price too low, leaving little room for negotiation on price.
Spare us the comments to the effect that we’ve oversimplified a very complex proposition. We already know that. Let’s hear your take on Zillow’s value to home sellers, and on the usefulness of Realtors in estimating market values in rapidly-changing markets.