Zillow

“But Zillow says my house is worth….”. That’s a fairly common statement, and it highlights how much the public trusts Zillow these days. I’ve written other posts comparing 10 actual appraisals with Zestimates, but today I want to show you a property in Sacramento that is over-valued on Zillow by more than 100%. While Zillow is a neat tool, let’s consider some of the important factors that go into making values accurate, and where things went wrong with this property.

Case Study: Let’s look at 5309 Broadway in Sacramento located in the Greenfair townhouse complex. You might be thinking, “C’mon, let’s cut Zillow some slack since this is a complex of only 45 units, and sales have been sparse for two years”. But Zillow sits at the adult table, and should have access to data from previous years. Moreover, notice above that Zillow actually shows the subject property sold in 2008 for $138,000, which gives Zillow a context to measure value.

The Zestimate: The Zestimate for this property is $321,679. The graph above shows all sales since 1998 in the Greenfair subdivision. As you can see, there have never been any sales above $300,000 in this complex – even during the height of the housing bubble in 2005. In this case Zillow is frankly wildly off since market value looks a whole lot closer to the red trend line.

Where did Zillow go wrong?

1. Choosing the Wrong Comps: The image to the right shows “nearby similar sales”, but these sales are single family detached homes, and NOT attached townhouses. When there are no recent sales in a townhouse subdivision, it doesn’t mean you should borrow from the single family detached market. Either you can use VERY old townhouse sales in the same subdivision, or maybe find a competitive complex somewhere in the market area. Again, it’s easy to cut Zillow some slack here since they might not know the units are attached, but even in that case the TINY lot size and history of sales should be given much stronger weight then.
2. Data Fail: Zillow clearly didn’t consider even its own data in this situation. Despite an understanding that this property sold in 2008 for $138,000, something in Zillow’s algorithm is obviously not crunching the numbers correctly since the market has not increased from $138K to $322K. Moreover, not considering a listing in the complex that expired at $186,000 this month is also a failure. When there are few recent sales, sometimes much older sales and expired listings can tell us about the market.
3. Wrong Neighborhood Boundaries: Zillow is considering single family detached homes in Elmhurst and other parts of Tahoe Park as you can see by the addresses in the “similar sales” image. These areas have far higher prices compared to the Greenfair subdivision. If you use the wrong neighborhood boundaries, there’s a good chance the value might be off-base too.
4. Problems with Less Data: When there is little data to consider, it looks to be a struggle for Zillow. To be fair, it’s relly not easy for humans to crunch numbers either when there are not many numbers to crunch. Yet data is available. It’s just a matter of seeing the numbers in their proper context.

One Buyer’s Reasons for Using Zillow: I asked a current buyer how she is using Zillow as she hunts for a home. I thought her response was interesting and insightful. What do you think?

At first, it helps me get an idea of overall neighborhood values, so it helps me know where to look or not look. Then, when we do look at specific houses, it gives me a general value of the house. I like the low-high range tool better than the “Zestimate” because it helps me get a feel for the overall values of a neighborhood. So, if a house is priced near or less than the low end, I figure it probably needs a lot of work, and if not, it might be a good deal. If it’s priced near or over the high end, I expect it to be in very good condition or have some kind of bonus features. Likewise, if we really like a house and it appears to be a good value according to Zillow, we’ll consider making an offer.

We also use Zillow to see a price and sale history of the house (our realtor can do this too, but it’s easy for us to do with Zillow rather than constantly calling her!). We can see when it first came on the market and various price changes, whether it’s a flip or not, and sometimes even if it was a rental.

One thing I don’t like is that it doesn’t have very accurate listing information. There are many houses on Metrolist and Redfin that aren’t on listed as “for sale” on Zillow. So, I find myself going back and forth between the three resources and our Realtor’s updates! If Zillow and Redfin merged, I’d be happy!

I look at Zillow as a range or estimate. I know that it doesn’t replace a person on the ground, but we can’t bring an appraiser with us to each house! 🙂 Zillow can’t see a smelly smoker’s house or a house full of old wall paper that needs to be torn down, or a crazy neighbor with three boats on the front lawn, or a dog that barks at all hours. It also can’t see a potential great neighbor with kids our kids’ ages, or a shade tree that’s perfect for a tire swing or tree house. It also doesn’t understand that I’m OVER granite countertops! Enough with the granite!

Zillow isn’t usually off by 100%, but cases like this are worth noting because they highlight some of the issues a “machine” can have when valuing a property.

Quick Advice:

1. Take Zillow with a grain of salt.
2. Don’t excuse Zillow when it’s wrong. If it’s off-base, call it what it is. You can look at Zillow’s own accuracy rates and be the judge whether this is reliable data or not.
3. Home owners, realize Zillow doesn’t know neighborhood boundaries, the condition of your home, all the same listings that are in MLS, and it may not even be comparing your house to the right type of property.
4. Agents, be sure to look up the Zestimate before listing presentations so you can be prepared to answer when your potential client says, “But Zillow says….”. Consider some of the positive reasons why consumers like Zillow (there are some for sure), but then talk about the things you know as an expert – neighborhood boundaries, the mood of the market, sales and listings in the immediate neighborhood, expired listings, how long it it taking to sell in the neighborhood, the direction of values, the condition of the house, and what buyers are willing to pay more for in the neighborhood.

I hope this was helpful.

Question: Any stories to share, or any other points you’d add?

If you liked this post, subscribe by email (or RSS). Thanks for being here.