Zillow’s Zestimate – Fool me once. Once is more then enough for one’s largest asset.

Zillow for me went from being an asset for quick and dirty home valuations to a liability. 

Okay if you are a real estate agent who is even slightly tech you’ve no doubt ran across Zillow and probably even tried to see if it would short cut the CMA process on homes.  For those who aren’t real estate agents, CMA stands for “Comparative Market Analysis” and it’s a process of using the MLS and knowledge of homes in the area to come up with a valuation of a home for a prospective seller to ultimately determine the highest sales price of the home in a reasonable period of time (say three months).

As a prior HouseValues subscriber who subscribed to the valuation requests, Zillow looked like the Holy Grail to a process that normally would take anywhere from an hour to three hours to generate a range of value that home might sell for and for all intents and purposes had the potential of reducing the period to 10 – 15 minutes, simply the time to transpose the data from Zillow into the HouseValues system.

It became obvious fairly quickly after running a few CMA’s that way that the valuations just didn’t feel very accurate.  Sure I knew the neighborhoods and in many cases had already viewed some of the homes which Zillow had called out as comparables and in reality the range of value that I entered wasn’t inaccurate, however it also didn’t feel accurate either.  It really came down to the fact that I didn’t have faith in the estimate of value coming out of Zillow.

What drove me to write this particular blog however is one of my past clients who purchased a home about three years back is now thinking of selling.  I would from time to time type in the address into Zillow, again with the idea that it would kick out a value that I could simply have on the top of my head should I talk to my client.  That number by the way was about $365,000 according to Zillow.  Recently she asked if I would be interested in listing it and I obviously I said I would.  The subject of price came up and without skipping a beat I suggested the home would probably list between $340,000 – $365,000.  (I didn’t want to err on the high side so I used the Zillow zestimate as the top end.)  If my CMA came back above $365,000 then we would obviously look at listing at the higher valuation.  I asked one of the agent on our team to work up a CMA to send to my client and much to my chagrin her CMA came in at $309,000.  Oh no!!!  What the heck is going on here.  Zillow is showing me $365,000 for a quick and dirty estimate of value and the agent on our team came back with $309,000.

What’s wrong with this picture.  Well what is obvious is that someone or something is wrong…  Guess what, I wanted to see the comparables.  I need to save face with my client (by the way I haven’t talked to my client yet…  This is a bit of my vent in advance of that), however the comparables look pretty darn accurate.  In fact based on recent solds and recent listings $309,000 looks dead on, anything more then $5000 – $10,000 above that is going to reduce the number of showings considerably and the likelihood of an offer beyond that.

I’ll probably get a little slack from my client (who has become a friend since she bought the house) when I tell her where my initial estimate of value came from, but it makes me sound a bit uneducated on the market by simply throwing out a figure without actually pulling the comparables before making the call.

I even decided to visit the Zillow site and see what it said about is valuation model and its quite telling if you actually read the fine print.  In Washington State the mean percentage from actual selling price is 5.7%.  That doesn’t sound too bad, however mean is different then average.  If you take all of the valuations that Zillow does and you compute the deviation from the selling price 5.7% sound good, however if the valuations that were more then 5.7% were a lot more then 5.7% then the average could in fact be a lot larger.  Another stat for Washington State was that 72% of the time the selling price was within 10% of the zestimate.  That means that in reality that’s a 20% potential spread 72% of the time.  On a $309,000 home that means that if they were accurate with their zestimate then for their estimate to be even in the ballpark it would have had to be a high of $339,900 and a low of $278,100.  This zestimate obviously fell into that 28% of the time that they are off more then 10% on valuation.  Assuming that we ultimately do sell the home for $309,000 that is a difference of in excess of 18% from the Zillow valuation.

My bottom line on Zillow is it’s a novelty.  If you want to for grins and giggles see what a computer comes up with on the valuation of your home then go for it, however I do not recommend that serious home buyers or sellers use Zillow without a second opinion from a knowledgeable agent.

Zillow is like the trusted friend who pulls a fast one….

Fool me once, shame on you!  Fool me twice, shame on me!  Zillow will not fool me twice.

Submit a Comment

Your email address will not be published. Required fields are marked *