How Good is Zillow's "Zestimate" of Home Value

73 Replies

While using the site Zillow to browse for multi family investment properties I have noticed what they call their "Zestimate." (their estimation of the home's value)  I would like to know how useful other investors have found this estimate to be?

Its within 20% about 80% of the time

@Phineas Howie I have heard that it really depends on your area. Not very reliable here in TX, sometimes it's on point but sometimes it's way off. I wouldn't trust it for getting an ARV.

Originally posted by @Mike Cumbie :

Its within 20% about 80% of the time

 And the other 20% of the time? Lol

I am looking in upstate NY,  Hudson Valley area. How do you generally estimate listed homes value quickly when Looking for good deals? 

Originally posted by @Tristan S. :

@Phineas Howie I have heard that it really depends on your area. Not very reliable here in TX, sometimes it's on point but sometimes it's way off. I wouldn't trust it for getting an ARV.

 Thanks for the advice, so how do you think can one find out how accurate it is in any given area?

In Texas, it is slightly less accurate than:

a Shaq free throw

a Tebow pass or

a Phil Mickelson tee shot on the 18th hole while leading the US Open

@Phineas Howie

Here in Pittsburgh, it depends. In 80% of single-family-home Zestimates, it's realistic to 10% of eventual buying costs about 60% of the time. However, when it's wrong the rest of the 40% of the time, we find it's as often wildly wrong as it is slightly wrong, that is, it is not slightly less accurate for the next 20%, and then just a bit less accurate for the next 10%, with 10% true outliers left over. No, it's within 10% of real sales price 60% of the time, and the other 40% it's inaccurate in a thoroughly unpredictable way. When it comes to multi-family investment properties, the Zestimate in the parts of Allegheny County that I've studied is less accurate than it is when it comes to single-family homes, to the tune of 50% on target within 10%, 50% inaccurate in the scattershot way I've described.


I would encourage you to think of the Zestimate as our outfit does, yet another tool in buying and selling properties, an independent statement of value that has outsized power over the decisions of casually-informed buyers and sellers, and even more importantly, a statement that can be manipulated to achieve your desired outcomes.

So my thinking goes like this: when YOU buy and sell as an investor, especially one making make-or-break first moves, you have to learn how to do your own comparative value assessments, and you have to learn how to do them using the same detailed techniques that the most experienced and successful local wholesalers and foreclosure buyers use, not agents who have the time to "pull comps" for free for every fresh-faced lookie-lookie with a preliminary loan approval letter in their pocket. To get within that hard 10% about 95% of the time, you also can't trust appraisers either hired by a bank or that you pay out of pocket.

This is how I do it for myself and how I back up my mentors in my own target area:

Physical inspections, county records, the MLS, Zillow.com, other real estate sites, accurate estimates of the true cost of repairs, all of these play a role. Your first step is Google Maps Street View. Your next step is a decent camera and YOUR OWN BOOTS on the ground. The next step is understanding the true cost of repairs in your target area, and that's a lot of work and knowledge. Then you have to master the local records system in that area. Then you study how the real estate websites, especially Zillow.com, handle sales pricing in the area. If you can, you make a contact with an agent who sets you up with access to the MLS in your target area. In Western Pennsylvania, there's a great computer system called "Matrix" that works with the West Penn Multi List. It can send you immediately email updates on every property listed for sale or reduced in price in a specific zip code. This is absolute gold for gaining the kind of in-depth sales knowledge of a target area I'm talking about.

If you want to really piss off a successful wholesaler at a foreclosure auction, ask her or him twice what he thinks the values of properties that are up for sale is, places that YOU haven't physically bothered to go to.

Depends on a lot factors.  Sometimes it's spot on and other times it's way off (in both directions).  BUT, it is an ok starting point and a reference.  If you base what you're doing solely off a Zillow estimate, that's obviously a bad idea, it's just another point of reference in your research.  You'll discover pretty quickly if it's accurate or not when you find several sources for reference.

@James K.

Thanks, so much for the sound advice. Really goes to so how much of an advantage it start somewhere you already know well like your home town or somewhere you have lived a long time.

More accurate than most real estate agent estimates, but probably not good in certain cases

@Phineas Howie

Oh, you're welcome, but of course I have an ulterior motive. I'll be hitting you up here in a couple years to get your informed opinion on deals in your target area. :)

Horseshoes...Hand Grenades...Zestimate.

Seriously, I read they are catching flack for the way they figure the estimates. Apparently, they are a crapshoot in many areas and buyers are relying too heavily on it when making an offer.

@Phineas Howie I live in lower Hudson Valley and I find the prices in my area to be within 5-10% of real value generally. Of course there are exceptions. It's probably a good tool to get a rough idea of value, but not more than that. In some other areas of the country it's almost worthless.

I usually use Zestimate as a starting point but take it with a massive grain of salt. NY is a disclosure state, which means that you can see how much houses have sold for without MLS access. After I narrow down my search, I start to pull comps from Zillow recently sold homes that come closest to the property I'm interested in. This will give you a better idea of the value of the property you are looking for.

I love my RE agent but they are not providing me with anything different that I cant pull off of Zillow or Realtor.com for sold properties.

If you print out the Zestimate, you could use it for toilet paper in a pinch, so not entirely useless :) Seriously, though, you can pull your own sold comps from Zillow and come up with your own estimate of value. If you want to be an investor, then pulling comps and coming up with market value is really a core competency that you should develop ... getting that value correct will directly affect your profits, so you want to learn to do it yourself from the ground up. Also, I'm assuming you are referring to 4 units or less when you say multi-family ... if it is 5 units or more, then that is commercial property and market value will be determined differently.

I don't use it and my job depends on comps :).

It is a simply a mathematical calculation.  It does not know any of the intangible.  It looks at the immediate area and makes a calculation.  It weighs the same property in distressed condition and in restored condition.  I have found it is most accurate in stable homogenous areas.  I live at the beach and every block closer to the ocean is worth 10%. In other words, a property 5 blocks away could be worth 50% more.  ZEstimate struggle with that as a concept.  An agent comp is probably your best bet.  If you really want to know have an appraisal.  It will cost a few hundred busks, but it may be worth it.  

All of the aforementioned comments seem to be in line with my understanding. Good for a "rough" estimate if you want a tolerable range for appreciation but not great to narrow in on what the true market value is. 

Moreover, to support many people's sentiments already, Zillow has also recently dealt with lawsuits for this very issue: Zillow faces lawsuit over 'Zestimate' tool

So everything on sites like zillow, redfin and sites like come from mls? or simly the larege majority of them?

@Phineas Howie ,

A lot of information does. The problem is that with the current algorithms there is only soo much that can be accounted for from a techie 18 states away. The value of an extra bath in a colonial in Brockport, NY is different than the value of an extra bath in a raised Ranch in Salt Lake City. The number of buyers vs Inventory in a stagnant area vs the numbers of buyers in a place that is building a new sports stadium effects market prices. Are there a lot of foreign investors or has an article been in the Washington Post recommending the area for investment? I firmly believe that someday the programming will be spot on, I just don't think it is now. I recently looked at one of the online sites that showed the comp properties consisting of 2 colonials and a split level, the subject property was a ranch.

I usually run an auto script from RPR, look at the estimated value in Reallist, the appraised value for taxes, then run my comps off the MLS and see where I am. I then make adjustments as needed based on market knowledge and what the buyers are looking for now. I still attend open houses because they are a good way to "hear" what people want. Even if I don't get that client, I do get market knowledge and feedback for a client.

My property in SC was appraised for ~215k

The Zillow Zestimate is ~195k

BTW when Zillow's CEO sold his house it went for 60% of the Zestimate...

at the end of the day the answer is, it depends. It can be very accurate or off by 40%. Its an algorithm so it all depends on the quality of the input data, such as how comparable the comparables really are that are being used. If there are a ton of cookie cutter houses as comparables and a lot of recent sales then you will probably get a more accurate representation of true value than if there are fewer sales and comps being used not really comps. 

Because of this, you should never use it as an actual input for your due diligence on value. Look up recent sales on redfin, have an agent pull comps, and if you want, have an appraisal done.

Free eBook from BiggerPockets!

Ultimate Beginner's Guide Book Cover

Join BiggerPockets and get The Ultimate Beginner's Guide to Real Estate Investing for FREE - read by more than 100,000 people - AND get exclusive real estate investing tips, tricks and techniques delivered straight to your inbox twice weekly!

  • Actionable advice for getting started,
  • Discover the 10 Most Lucrative Real Estate Niches,
  • Learn how to get started with or without money,
  • Explore Real-Life Strategies for Building Wealth,
  • And a LOT more.

Lock We hate spam just as much as you

Create Lasting Wealth Through Real Estate

Join the millions of people achieving financial freedom through the power of real estate investing

Start here