“Tsuru” over in Seattle Bubble comments, asked me for a comparison of Zillow Zestimates vs. Closed Sale Prices in the current market. To be sure the Zestimate isn’t picking up the recent sale, I’m using the latest 50 or so sales recorded in the mls for King County in the last few days. 42 are single family homes and 8 are condos. I’m only showing the data for the single family homes, but thought you’d like to know the breakdown of the sales for the last few days.
I think I saw David G. at Zillow and someone from Cyberhomes going at it recently, so let’s throw Cyberhomes in the mix too. As usual, I am posting this as the results come in…so I have no idea how it is going to turn out. Let the best “man” win 🙂
Also of particular interest are the number of sales that are Short Sales and Bank Owned or other “stressed” sales, many, and very few of those indicated so in the Public view vs. Agent fields.
Sold Price Zestimate Cyberhomes
$262,000 SS* $275,000 $292,552
$365,000 $301,500 $398,192
$287,000 BO* $311,900 no result
$530,000 CO* $743,000 $695,991
$140,000 BO* $186,957 $196,698
$210,000 SS* $269,500 $274,417
$282,500 $320.000 $281,461
$285,000 $276,000 $276,134
$347,500 $334,000 $347,910
$480,000 $422,500 $483,891
$550,000 ES* $705,500 $688,842
$565,000 NC none none
$652,500 ” N/A” $661,320
$190,000 BO “no result” $234,017
$269,950 SS* $285,400 $282,102
$279,900 $413,000 $272,349
$517,000 BO $682,500 “$0-Foreclosure”
$450,000 $454,500 $518,982
$176,000 BO $312,500 “0-Foreclosure”
$267,999 NC n/a n/a
$740,000 SS* $937,000 $794,218
$325,000 CG $547,000 $567,171
$420,000 $410,000 $402,384
$451,050 CR $637,000 $559,188
$850,000 NC n/a n/a
$835,000 $802,500 $811,305
$915,000 $831,500 $875,266
$850,000 NC n/a n/a
$370,000 $367,000 $428,766
$636,500 $702,000 $676,200
$650,000 SS * $707,500 $662,000
$360,000 $347,500 $376,152
$475,000 BO* $584,157 $585,199
$292,500 NC n/a n/a
$305,000 NC n/a n/a
$309,950 NC n/a n/a
$373,000 $342,000 $352,252
$386,000 NC n/a n/a
$389,950 NC n/a n/a
$400,000 $391,000 $392,337
$577,000 TR $467,500 $405,413
$416,000 $468,500 $484,506
*disclosed in Agent Remarks or owner field, but NOT in public remarks
SS = Short Sale. CO = Corporate Owned, CR = Corporate Relocation, BO = Bank Owned, ES = Estate Sale, NC = New Construction, CG = Completely Gutted, TR = Totally Remodeled
Geographically, most of the short sales, all except one, are South. The first sales are in Federal Way, Auburn, then Burien, Kent, South Seattle, over Mercer Island, Eastside, Bothell, North Seattle on the Green Lake/Greenwood side, then North Seattle up through Shoreline. That is how the mls code numbers run from 100 through 715.
Wow! That is a lot of work and I certainly wasn’t expecting an answer as comprehesive as this. Thanks Ardell!
Interesting, Ardell.
General trend – closing lower than zestimate in the south, higher than zestimate in the north.
How recent is the data? running through mid-late Dec?
How is January shaping up?
Seems like a lot of work to determine Zillow isn’t accurate, when they provide the data themselves.
http://www.zillow.com/howto/DataCoverageZestimateAccuracy.htm
Ardell
great post.
on first glance it looks like Cyberhomes is closer to market value.
if there are any statistic people out there, i am sure they can calculate the variance for both data sets to see which is most accurate.
what jumped out at me is the calculated values both zillow and cyberhomes were generally above actual sales price.
this makes sense – the comp sales used in the valuations are “historical” and in a down market, do not keep up with the decline in value. it is the same principle that applies in a hot market. in a hot market, prices are going up and so the sales data 3 to 4 months prior does does not reflect the continuning increase in values.
A lot of the lower price ones are short sales and bank owned, and you would expect Zillow to be above them because they don’t assume gross incompetence in the marketing of property when they come up with a Zestimate.
Hey Ardell,
Thanks for doing the leg-work on this! If only you could share the original list prices 😉
Kary is correct – you can’t learn anything from the BO and SS transactions – just like a Realtor would never list at those prices, we are not trying to estimate distressed sales values.
It would be great to have someone aggregate this list for us – if you e-mail me the spreadsheet, I’ll try to get to it today.
Arn –
Obviously that challenge also impacts appraisers and Realtors’ CMA’s. AVM’s typically don’t actually pick a small set of comps the way you’re thinking they do – rather, all local sales are analyzed and then sliced and diced to dissect trends by property type etc. Once the market is headed down, the trend actually becomes predictable and so the most challenging estimate (for anyone) is the one just after the market has peaked. In my experience, the fact that AVM’s analyze such a vast amount of data ands have no emotional issues with recognizing a downturn usually means they identify the peak before the market does.
Great piece, Ardell. I understand David G.’s comment that his data is historical. However, my question is, what value is this data? I’ve always questioned it, in any market, because the Zestimates are not accurate. The consumer needs accuracy, not transparency if the transparency is not accurate.
As David G. says: “all local sales are analyzed and then sliced and diced to dissect trends by property type etc. ” Analyzing all sales to determine the the value of a home is part of the problem. One size does not fit all.
In fact, careful analysis of only those sales most comparable to a home, whether it’s location, style, size or number of bedrooms, is the best way to determine a more accurate value of a home. Viewing the home in real time to see what the neighborhood is like, the street appeal of the home, the condition/upgrades of the home, the functionality of the floor plan, and the positives and negatives of each home are critical pieces of information when determining an accurate value.
I live in Kirkland, and as I know you do Ardell, and Kirkland is a prime example of this. The homes vary so much in price, even on one block, because we have older bungalows, mid-century ramblers and McMansions all on that block. In addition, there are micro-markets within a few short blocks of each other. I would never compare my home to a home in an area just a few blocks away, because I know it is considered to be a different price point.
As David G. says:
“Obviously that challenge also impacts appraisers and Realtors’ CMA’s. AVM’s typically don’t actually pick a small set of comps the way you’re thinking they do.”
So to get an accurate opinion of value, only a small number of homes, true comparables, should be considered.
Thank you, Tsuru. Fact is, I’m doing stats all the time from all different angles for my own reasons. While the post focuses on Zillow & Cyberhomes, I am really looking at many other things, as David points out.
For me there are always two questions:
1) Should my buyer clients buy or wait, and what are we waiting for. Are buyers who are buying making wise choices. Are the things that are selling, things I would recommend that my clients buy…or not. There are two properties over in Lakeview in Kirkland that are screaming deals, compared to any of the sales above, but who has a million dollars to buy them, even if they are screaming deals? To know if something IS a screaming deal, I can’t only look at my normal methods of valuation, I have to continually test that against the REAL market, which is what IS selling, not what isn’t selling or what sold 4-6 months ago.
2) For my seller clients who are NOT in the type of stress imposed by SS or BO or CR, etc…What are the odds they CAN sell at all, if they are in the area where 75% of the sales are stress sales? If there is 8 months of inventory BUT 1/2 of the sales are stress sales, that jacks up the absorption rate to 16 months of inventory for the seller not in a stressed situation. It’s like two different marketplaces, side by side. Like Norstrom’s and Nordstrom’s Rack being next door to each other.
I’m looking at how much are people getting paid for improvements? What is the discount for original condition 1969 or a 1980s kitchen. Was that totally gutted house the best deal…or not?
Kary…I bother to do the work because I ALWAYS bother to do the work…I’m almost always working, even if I only show in the post a portion of that work. I don’t blog vs. work…I blog while I work…and I don’t blog everything I’m getting out of the research.
I accompanied an agent showing a pricey condo the other night. The clients asked me a question. Off the top of my head I said: There are 17 SFH for sale in this neighborhood and no pendings. There are 37 condos and only two are pending. There are no short sales or bank owned properties, etc… Sure people look at me like I have a computer in my brain, but isn’t that what we DO for a living? Give people advice about buying this property based on our experience and knowledge? I had no idea I was going to be in that neighborhood when I did the research.
I can’t stand agents who view property and go “oooh, I LIKE this”. Who gives a RA what an agent likes? Who needs an agent to tell them the neighbor has junky cars out front. If we’re not going to always work to be “on top of the market”…where it’s been and where it’s going…who needs us?
” I would never compare my home to a home in an area just a few blocks away, because I know it is considered to be a different price point. ”
Very important question, Debra. Considered different by whom? Studying what is selling and for how much is VERY important to agents, as we don’t get to tell people that they HAVE TO pay more once they cross this street. In changing markets, the premiums for snob neighborhoods is often the first to erode. Look at West of Market vs. East of Market and how the buying public changed agent’s perception of value there in the last few years.
If we are not looking at the general trends, and the change in what buyers will and won’t pay for, we are imposing old fashioned valuation techniques that may or may not be currently relevant.
Bottom line is we are not the God’s of valuation IF we do not constantly look at stats and trends. We don’t dictate value…we study buyer patterns, and to study buyer patterns, you have to be looking at everything…not just the neighbors. Look at how many houses are not selling…because hindsight is being used more often than living in the present and future.
The only way to value property today, is to be inside the heads of ALL buyers who are willing to buy…as much as humanly possible.
David G.
I don’t do spreadsheets…I always write “on blog”. The only really odd Zillow Zestimate was the one I DMed you on Twitter.
Original asking price was irrelevant, for those asking that question.
Here’s what WAS relevant. Heck…I may as well do a follow up post 🙂
Before I move to the follow up post, this post is like Trulia Voices. Someone asked a question and I gave the most comprehensive answer to that person’s question. If someone asks the “wrong” question, like what were the original list prices, which is irrelevant in this data set, I answer the question they should have asked…which I will do now in the next post.
Hi Ardell,
I agree with your thoughts. What you mention in #9 is a huge part of our job. I, too, follow stats like a hawk and report on them on my blogs and to my clients. It is critical that we do that for our clients, both buyers and sellers. They rely on us to tell them the truth about the market, its performance and how we see value. My opinions are based on the stats.
What I wrote above is another part of the puzzle and that is determining value for homes for either our buyers or our sellers. Ardell said: ” we don’t get to tell people that they HAVE TO pay more once they cross this street.” I agree with that sentiment. I don’t tell people what they HAVE TO pay, only about what I believe is more of an accurate comparable and how the market is performing (stats). I give buyers my opinion as to how to proceed with an offer, based on where they tell me they want to end up with pricing, timing, etc for a house. I advise people and then follow their lead as to how they want to proceed. It is always their decision.
Stats play a key role in any market. I’ve been following and reporting on them, giving this information to my clients long before I started blogging.
My comments above #7,were more about the inaccuracy of Zestimates, not about stats, and how they do not give an accurate picture of the marketplace.
The simple answer to the question would be to point out that it’s impossible for a site like Zillow to be terribly accurate because they don’t know anything about the condition of the property. I’ve heard appraisers indicate they are only accurate to plus or minus 5%, and that’s after they go in and pick comps that are the best comps they can find (which is something different than what the automated sites do).
The link I posted for Zillow is nice because it’s basically Zillow publishing their own shortcomings. They could sit on that data and hide it, but they don’t.
Thanks for the eye-opening data, Ardell. I thought Zestimates were basically useless in this market, and that most properties would sell for well under the Zestimate. But your data shows that that’s not necessarily the case.
Hi Debra,
For the record, I agree with you about as often as I disagree with Kary 🙂 You and I in a room together would probably be the most any buyer or seller could hope for from ‘a” professional 🙂 I love it when we put our heads together.
In my follow up post, the asking price of a home that sold for $740,000 was originally $1.4 million. The general public having access to Zillow and Cyberhomes and the Assessed Value is SO very important!
At least they are getting a head’s up that this owner might be asking twice what it is worth. Can you imagine if they just “negotiated” a substantial amount off asking price? That’s why I say “negotiation” is the lazy agent’s mantra.
sampai,
The Zestimate, coupled with other factors, is of value, but not merely in one direction or another.
The newer the property, the less value you place on things like Zesitmates and Assessed Values, BTW. The newer it is, the more likely everything will be inflated. The older it is the more likely everything will be looking less than it should. I’m talking older as in 1910. In the 60s and 70s it’s all about the condition of the comps, vs. the condition of the subject property, as in that last sale at $417,000. Original condition vs. the comps having more updates since the 60s.
Condo Conversions would fall in the newer category, even if the building itself is not newer.
Zestimates and other valuations available to agents and the public are of value…they are one of the things you look at to get to the end of the job at hand.
Zillow is further off the longer it’s been since the property was sold. If the owner has owned for 20 years it’s more likely to be incorrect than if it were sold only 2 years ago.
Kary,
That is only true if there are few other sales in the neighborhood, which is not often the case until recently. Zillow isn’t stuck with some 20 year old valuation unless it’s a cabin all by itself in the woods.
Ardell wrote: “The newer the property, the less value you place on things like Zesitmates and Assessed Values, ”
Was this a joke? I can’t believe any professional agent would place any weight at all on a Zestimate when doing their analysis. I guess there’s a reason we disagree a lot.
Maybe appraisers should pick this up. They could do comparable sales, replacement cost, income analysis and comparison to Zillow! 😀
“That is only true if there are few other sales in the neighborhood, which is not often the case until recently. Zillow isn’t stuck with some 20 year old valuation unless it’s a cabin all by itself in the woods.”
Wrong, but I don’t have time to go through a bunch of sales to show that Zillow data is wrong and how it’s wrong.
Hey Ardell – I wasn’t asking – just being my smart-ass self
Did you get my DM’s about that other property?
Kary,
Most days I think you don’t understand the value of technology and the internet. It is up to us to help make it better, not to tear it down.
How did appraisers help people during the bubble? How did appraisers protect the banks and the buyers from over investing? Answer…they didn’t.
So if Average Joe Buyer can see an asking price of $1.4M and a Zestimate of $937,000 and an assessed value of $800,000…aren’t they better served than simply negotiating against the $1.4M asking price?
The internet is not about you and me as professionals. It is about checks and balances and a bit of empowerment for Average Joe Buyer.
Hi David,
No I didn’t go over to Twitter today. Will check.
“Wrong, but I don’t have time to go through a bunch of sales to show that Zillow data is wrong and how it’s wrong.”
In the many months we have “spoken” you never seem to have the time to do your own data to back up your objections. Just links to other people’s writings whose opinions you obviously value more than mine.
Maybe some day you will back up your objections with facts.
Dontcha know Ardell, they’re backed up with Truthiness!
http://en.wikipedia.org/wiki/Truthiness
deejayoh,
I LOVE it! I really like Colbert better than Stewart, but not as much as Maher. Anywhoo, I am SO full of Truthiness, it’s not funny! I love Truthiness! Whenever the answer is not an exact science, you have to pitch at it from all directions. Zillow is one of the pitches. The answer lies at the bottom of everyone’s “truthiness”.
Thanks for the chuckle!
Ardell,
I am not sure why think Kari needs to prove Zillow is inaccurate for you. Zillow provides the proof and freely states the inaccuracy of their WA data on their website for the public to read. Stan, VP Data and Analytics, will tell you the same. I have had a conversation with Stan about it at one of their company open houses. Junk in junk out. Zillow is a nice toy in the toybox.
Assessed values and property taxes are based on government fiscal cycles and the need the generate revenue to pay for government services and operations. It is historical data. There is little correlation to FMV
Michael,
I don’t think you or Kary understand the general public’s need for info beyond, “don’t worry your pretty little head about it…we’ll tell you what it’s worth”.
Help make things better…don’t tear them down. People want more info…and day after day many try to give it to them and improve on it.
Kary
What does:
>>gross incompetence in the marketing of property ..<<
mean?
Ardell wrote: “I don’t think you or Kary understand the general public’s need for info beyond, “don’t worry your pretty little head about it…we’ll tell you what it’s worth
Kevin, by “gross incompetence” I primarily mean not providing timely answers to offers. That greatly reduces the pool of buyers, and thus the potential price. For REOs you could also include shoddy marketing, such as houses that are not in the greatest of condition, to put it mildly. Zillow has no way of knowing that the roof and yard are covered with leaves, that there’s no heat inside or running water and that the place is generally filthy. Not exactly the conditions that get top dollar.
Kary,
The most up to date is the most useful. I will always do my own stats and never rely on anyone else’s data. We are in times of exceptional difficulty in determining the number of ready, willing an ABLE buyers in the marketplace. Supply and demand is key, and demand coupled with ability to purchase is changing every day. Every time I speak with a lender I hear about another road block to buyers who would buy but can’t.
Every day I see properties not selling at $400,000 to $600,000 under assessed value, due to financing issues, and cramming down prices that will ultimately impact all price levels.
While many are tracking the continued diminishing of volume, we are watching and listening and studying from the front row to determine the WHY. No amount of CURRENT information is “too much”. Looking back is not useful in this market…looking forward and paying very close attention to every snafu in every transaction is key.
You decide what is “more useful” for you and your clients. I’ll decide what is “more useful” for me and mine. You really don’t have to agree with me..but you really have to stop constantly disagreeing with me as if you hope to change me into you…that is not something I aspire to be.
I don’t see Zestimates as serving any purpose for real estate professionals, so I don’t see being up to date meaning anything.
BTW, Zestimates are based in part on assessed values, so by relying on both your doubling up.
Kary…take the professional out of the picture. How does a buyer begin? Until the answer is they CHOOSE to have a professional vs. they MUST have one because no info available to them is of any use whatsoever…we all have to fight to get people better info.
You make it sound like any old professional is better than anything else. I’m sure you know that is not true.
Ardell,
From Ardell: I don’t think you or Kary understand the general public’s need for info beyond, “don’t worry your pretty little head about it…we’ll tell you what it’s worth
Zillow is a good source of information, but the Zestimate is not, and most likely never will be. There’s a limit to what you can do with computer analysis of public records when it comes to valuing property.
The problem with much of the information that is kicked around on the internet is that consumers often do not know how to interpret it or they rely too heavily on that data as fact. As a lender, I see this daily as people still think 30 year rates are around 4.5%. Or they go on sites like Bankrate and think their “bank rate average” is really accurate when in fact it typically is about .5% too low and they are confused when the quotes they get are no where near what they read on the internet.
I can’t tell you how many times I get borrowers complaining that the appraised value is lower than their Zestimate or I check zillow first and I get the appraisal back $100k higher.
I truly believe that consumers are sometimes given too much information without context or professional interpretation and it often leads them to make misinformed decisions. Most people do not read the fine print…
Russ wrote: “The problem with much of the information that is kicked around on the internet is that consumers often do not know how to interpret it or they rely too heavily on that data as fact.”
A good example of that would be the emphasis on price. The newspapers give us at least three difference sources of information on price for each and every month, spread out over a period of months.
Newsflash! Price probably hasn’t been the most important piece of data for at least a year (if it ever was). Volume took a huge hit at the end or 07, beginning of 08, and another huge hit at the end of 08. Your house is only worth $XXX,XXX.xx if someone is willing to pay that amount for it, and with the current volume levels that is rather problematic, especially for the less desirable properties.
Here’s a challenge to whoever. Find a consumer site or news article that gives significant data on volume. Oh, and the news stories from the beginning of the month that erroneously blamed the low volume for December on the snowfall don’t count. I’m familiar with them.
Edit: This piece at least takes a stab at it, but I think the charts are somewhat confusing, to the point of hiding the information, and there isn’t a heck of a lot of commentary that would perhaps help explain the charts as to volume.
http://seattlebubble.com/blog/2009/01/20/fourth-quarter-sales-in-the-gutter-delistings-stable/
“The problem with much of the information that is kicked around on the internet is that consumers often do not know how to interpret it or they rely too heavily on that data as fact.”
This fear is largely unfounded and certainly not peculiar to information “on the internet.” It’s a simple fact of adult life that we all have to filter everything we read and everything we hear. I’m actually quite surprised that I have to remind you of “Caveat Emptor”.
I just logged into twitter. The first tweet about Zillow is from a consumer. It says … “Another way to waste hours of your time-check out Zillow.com.It will tell you approx how much your and all your neighbor’s houses are worth.”
Note: “approx.” Consumers get it.
Zestimates serve as a reminder to buyers and sellers to properly query the market value of a home. If you can’t understand the value of that reminder then you probably have more dollars than sense. 😉
David G:
I should clarify that I am not bashing Zillow. I have wasted plenty of hours playing around with Zestimates. It is a great tool. My point is just that consumers often do not get it. I say this as a professional who deals with them on a daily basis. Even Zillow’s message boards are filled with people complaining about inaccuracy of the Zestimates because they treat it as FACT instead of just a data point to consider. The issue is that there is no CONTEXT and consumers generally do not read the fine print. It says it on a big website such as Zillow so it must be true is the attitude of probably 80% of the public.
The internet is wonderful and many see it as this disintermediation device that will cut out the middle man to their own detriment. Why do I need a Realtor? Zillow tells me what the home is worth. Why do I need a loan officer, Bankrate tells me the rate average. Why do I need an attorney, I can get my divorce papers or real estate closing on line. If only it were that simple.
I wish someone would pay me what Zillow says my house is worth…
“I wish someone would pay me what Zillow says my house is worth…”
Sounds like a blog post title! 🙂
Russ –
How is this disintermediation argument relevant here? Most (all?) consumers do not think that Zillow is going to replace their Realtor. And the few people that do … those people believe their goldfish could replace a Realtor; they didn’t need Zillow’s help. We certainly don’t believe that our site replaces a Realtor. Do you? These conspiracy theories are holding RE pro’s back from productively embracing effective marketing technologies. It’s bad for business.
We put an estimate on more than 60 million homes. Many people would like that number to be different. That should not surprise you. It should also not surprise you that some of those owners write to us quite passionately about how they would like that number to be different. But none of that suggests that the average man on the street does not understand that a Zestimate is an esttimate. I speak with the angry sellers and owners every day – they would absolutely LOVE for us to give their Zestimate a boost – but they simultaneously do get the fact that it’s an estimate. People are smarter than you’re giving them credit for.
Just saw this in Twitter: “Zillow keeps emailing to let me know the value of my house is decreasing…um, thanks for the updates…I guess.” Biz Stone “the co-founder of Twitter” – Berkeley CA
David:
The point is that consumers see all this information on the internet (not just zillow by the way) and think it is FACT. Again, I love the website. I use it daily as a gut check when I get a client who calls and says their house is worth X. I am sure there are plenty of consumers who understand that it is just a tool and one data point, but I come across a lot more who take it as the gospel. In fact, I am not necessarily even singling out Zillow, but really directing my comments to information websites in general.
I think you are underestimating how much consumers rely on very visible websites like Zillow, Bank Rate, and the mainstream media in general and assume the information is 100% accurate. Half my clients have MBAs & Law Degrees from the very very best graduate schools in the country and I am constantly having to explain that just because they saw an article in USA Today about 4.5% interest rates doesn’t mean it is absolutely true.
I brought up disintermediation because I think consumers have access to more information than they ever have before and it gives them a certain confidence to try to cut out the middle man even when they may be making that decision on bad or incomplete information to their own detriment. Many information sites like Zillow often times indirectly promote that mentality among consumers.
I am sure as time goes on you guys will figure out how to refine your algorithm and make the model more accurate. Just don’t underestimate how much weight people give it though.
Pingback: Bridget Magnus » But Zillow Said!
Russ, people believe a lot of things they see and read that simply are not credible. 60 Minutes comes to mind. Very popular show, but hardly accurate reporting. Campaign ads. Presumably they’re effective, right?
I suspect whether people believe Zestimates is largely dependent on whether they like the result.
SS = Short Sale. CO = Corporate Owned, CR = Corporate Relocation, BO = Bank Owned, ES = Estate Sale, NC = New Construction, CG = Completely Gutted, TR = Totally Remodeled
These are what I would call unzillowables, factors which Zillow is not privy to, which influence the value of a home.
This post also shows the value of having access to MLS data.
Some questions to David G:
1. What exactly is the value of knowing the median error rate for an area? 2. How does it help me when I look at a specific home on Zillow?
3. Should I take comfort that the home is likely near the median?
4. Given the median is merely a middle number, isn’t it possible the particular home is well above the median error rate?
Joseph, the median error rate is different than a house being near the median value. Check out the link I posted above:
http://www.zillow.com/howto/DataCoverageZestimateAccuracy.htm
Per that the median error rate for Seattle is 7.5%. That means that on 11 $500,000 houses, 5 of them will be valued by Zillow at either over $537,500, or under $462,500, one of them would be valued at either of those two numbers, and the remaining 5 would be within that range.
That site also shows you that for Seattle, 17% of the properties will be outside of 20%. That means 17% of $500,000 houses would be valued at over $600,000 or under $400,000.
Hi Joe 😉 Apparently Kary is not familiar with the “Joe goading David G.” conversations going on around the net for more than a year now. LOL.
However, given the state of the market, and the specific values in this post, this statement doesn’t seem to hold true ” isn’t it possible the particular home is well above the median error rate?”
Seems more likely to be under vs. above, based on the recent “regular” sales in this post.
Joe,
Has anyone ever asked you this? Is the Zestimate of your home more or less than what you think your home would sell for today?
Thanks Kary.
So, if the median error rate in Seattle is 7.5% what is the value of 7408 52nd Ave NE Seattle WA 98115? The zestimate is $629,500
I don’t know Joe!
That said, being “under the median error rate” as I would interpret the phrase would mean that the house was within the median range. It wouldn’t mean it was over or under valued by Zillow–it could be either.
That’s the problem with the stats that Zillow gives out. They don’t tell you how many are over and how many are under.
Joe, you probably should edit that address out within the next 10 minutes while you can. If you miss that time limit, perhaps Ardell can do it.
And BTW, without viewing the property it would be difficult if not impossible to accurately give an estimate for that particular property. If I did that I’d be doing exactly what Zillow does–assuming you can accurately value property over the Internet.
Not sure why Joe picked that particular house, but why should I edit Joe’s comment, Kary?
@Ardell What I mean when I say “well above the median error rate’ is the true home value is off by more than the median error rate. So, if the median error rate is 7.5%, if a house’s true value is off by more than 7.5%, lets say 25%, this can be either above or below the zestimate. Thus, a house with a zestimate of 100K that sells for 75K or 125K is, either way, off by 25%, which is ABOVE the median error rate of 7.5%
The point is– knowing the median error rate is of absolutely no value when looking at a particular home. Unless you know the house value (in which case Zillow is unnecessary), knowing the median error rate provides no comfort, guidance or assurance as to how far off the zestimate to THAT home is.
“The point is knowing the median error rate is of absolutely no value when looking at a particular home. ”
Maybe people are just smarter in Seattle 🙂 I’m looking at the sales in this post and it’s pretty obvious that the people who bought them knew which way to move away from the Zestimate, and by how much.
The $416,000 Sale Price vs. $468,000 Zestimate was right off the main drag on one of those streets to nowhere, and it needed total updating including the 1960s silver colored aluminum windows.
The $373,000 Sale Price vs. $342,000 Zestimate was in a neighborhood where most of the houses (the comps) have freeway noise. But the subject property was in a WOW! best location with no road noise and backed up to the park/golf course.
Looks like the people can take the Zesitmate and add and subtract based on the home’s particulars fairly well around here. Like I said, maybe Seattle people are just smart enough to use all their senses. They can see old windows, they can hear freeway noise, and they can smell cat piss. The don’t expect Zillow to do that for them.
Joe – the Zestimate value is an estimate. It’s not an appraisal. To find out the value of that home …
… if you are selling this home, contact Ardell or a professional from this list: http://www.zillow.com/directory/seatlle/real-estate-agents-sellers-agent/
… if you’re the owner and need an appraisal for financial reasons, contact a professional from here: http://www.zillow.com/directory/seatlle/real-estate-services-appraisal/
As to the rest of your post … please refer to the first one thousand times I answered these questions over the past 3 years. The answers haven’t changed and I doubt I’m not the only one that’s tired of your hijacking otherwise productive conversations like this. If anyone else feels like they would genuinely benefit from having me go over this with Joe for the one thousand and one’th time, please speak up and explain your issue so I can address it.
@Ardell OK smarty pants 🙂 so how far off is the zestimate to 7408 52nd Ave NE Seattle WA 98115? The zestimate is $629,500.
“I’m looking at the sales in this post and it’s pretty obvious that the people who bought them knew which way to move away from the Zestimate, and by how much.’
Hmm.. how did you know they all used the zestimate? You really are smart 🙂 🙂
Ardell wrote: “Maybe people are just smarter in Seattle 🙂 I’m looking at the sales in this post and it’s pretty obvious that the people who bought them knew which way to move away from the Zestimate, and by how much.”
That’s crazy because it’s starting with a number that could be high or low, and is just as likely as not to be off by more than 7.5% either direction.
That makes about as much sense as going in and offering 90% of the list price, without knowing anything about the house.
And quite frankly, of the 20 or so examples you gave in the original post, I’d be surprised if more than 2 of them knew what the Zestimate was when they made their offer.
As to 55 I wrote that before checking and assumed it was either his property or an active listing.
Joe,
LOL..well let’s say IF they used the Zestimate and then added and subtracted as I indicated, they would have been smart to have used the Zesitmate as a tool in the process 🙂
Using someone’s actual address around here, makes Kary nervous 🙂 Did you pull that out of thin air, or do you know the people who live there? Kary hasn’t answered yet as to why he wants me to take the address out of your comment.
Kary,
Joe lives in Bucks County, PA…last I heard, in an area where I once lived and worked.
Kary…what number do you want people to “start” with? What number do you “start” with? You complain no matter which number we use, most times.
I tried using your logic in response to one of your recent comments, and could not find 3 sold in most any neighborhood, or like kind properties. I do see agents trying to use comps in the neighborhood that are more than 6 months old for that reason. Zillow’s Zestimate may be as good as “old” comps…no? They don’t exactly pull the number out of thin air. They do have a basis for the number, just like you have a basis for yours.
Most agents “start with” the asking price, and then yell about how well they negotiated…even Redfin does that. I don’t see why the Zestimate or
Assesed Value is of no consequence, when most agents use the asking price and what the seller will accept as a guide. That’s not any more accurate, for sure! Worse even.
I certainly don’t start with Zillow or a list price. At best the list price provides a ceiling when there are several listed at $XXX,XXX, but none are selling at that price.
I’m not really sure why you need a starting place. If you pick an arbitrary staring place, that will affect your ending place too.
Love you Ardell.
IMO, stating the margin of error is a red herring used by Zillow to give the public the perception the zestimate is more valuable than it actually is. But knowing a general median error rate does not give me any assurances as to the actual value of a specific home. I have NO WAY of using that median error rate as a tool to tell me whether the error rate on this house’s zestimate is above or below the median, or by how much. It’s the same problem as using a general average to reach a conclusion on a specific case. Heck, if I relied on a general average, I’d have to say a human being has one breast and one testicle.
ROTFLMAO even though I don’t see how a general average gets to to one breast and one testicle LOL! Well, one breast maybe 🙂
Love you too, Joe. If we don’t see David, we’ll have to assume he has a new policy of ignoring you 🙂 He could just be out of pocket.
So, Kary, if there are NO recent sales in the neighborhood…where do you start?
Expand the neighborhood or go to properties that are not quite as close as what you’d like. Do you think there are properties that appraiser can’t issue an opinion of value on?
Kary,
Yes..there are properties that an appraiser cannot accurately determine value on. Very much yes, on that one. Not enough sales in the neighborhod and he wouldn’t go out of the neighborhood, so he just appraised it for $100,000 less than sale price, since there were no like kind comps within 6 months nearby. The buyer switched lenders. We’ll see what happens. Appraiser decided to deduct 24% from his valuation to adjust for market conditions. Odd, but true. (not my transaction – I know the agent who is in the middle of that one)
I’m seeing properties on market for $600,000 less than Original Asking Price, with Original Asking Price being within the range of “the comps”. This market is hard to call no matter where you start… Anyone suggesting otherwise isn’t paying attention.
Like I been saying…it’ ain’t about what’s selling. It’s about what isn’t selling or isn’t closing once sold.
And YES, Kary…financing has changed since September. I know of at least a few buyers who had approvals in September, who can no longer get financing, even though their credentials have not changed.
Look at Biliruben’s story…my God, what the hey! Seems even when you think everything is in place, something goes sideways.
I think it’s called “We can’t win for trying”. Have to go look that old saying up. I’ve never seen anything quite like this.
Reposting David G.’s comment #58. It was trapped in the spam filter because of the links and I didn’t see David’s request until I looked over at his note in Twitter.
You can also see it up at comment #58
Joe – the Zestimate value is an estimate. It’s not an appraisal. To find out the value of that home …
… if you are selling this home, contact Ardell or a professional from this list: http://www.zillow.com/directory/seatlle/real-estate-agents-sellers-agent/
… if you’re the owner and need an appraisal for financial reasons, contact a professional from here: http://www.zillow.com/directory/seatlle/real-estate-services-appraisal/
As to the rest of your post … please refer to the first one thousand times I answered these questions over the past 3 years. The answers haven’t changed and I doubt I’m not the only one that’s tired of your hijacking otherwise productive conversations like this. If anyone else feels like they would genuinely benefit from having me go over this with Joe for the one thousand and one’th time, please speak up and explain your issue so I can address it.
Regarding the credibility of Zillow in the minds of the average buyer or seller, I thought you all might appreciate a viewpoint from the appraisal side of things. After all, we certainly do hear from those homeowners who vehemently believe our valuation is low, and we certainly hear why. And the interesting thing is, they never cite Zestimates. They will occasionally use Zillow or similar sites as comparable generators, but I’ve never had anyone simply say “Zillow says it’s worth this, so how can your value be different?”
Although I run into very few homeowners who trust Zillow’s Zestimates, a great number seem to use Zillow. Mostly they seem skeptical about it, but they use it because they have so few sources of information. Actually, one of the top questions I’m asked, after “how much is it worth,” of course, is “do you think Zillow is any good?” People just seem starved for information.
So I guess the question is, is unreliable information better than no information? As long as people understand just how unreliable it is and its limitations, and most do, I think the answer is yes. That said, I don’t think it’s the Zestimate that is useful. What’s useful is that it gives people who would othewise feel lost in the woods a place to start, and it gets them to start thinking about comparable properties and how to begin establishing a value.
Comment #72. I totally agree.
Re 17 and 24, I went back to get some data. I too picked recent sales, finding 27 SFR in King County and picking every third one to get a sample of 9. I had to throw some of the nine out because they were: (1) Sold in 12/08–too recent; (2) New construction; or (3) The most recent prior sale was a foreclosure sale (and that one was goofy, where it sold at foreclosure for over 800k a few months ago, Zillow thought it was worth over 1 million, and it sold for about $500k). The sold prices ranged from 310k to 1100k, with a mean of 502k.
Anyway, here are the results, showing the percentage Zillow was off (a negative number meaning Zillow was low), and the number of years since the last transfer for value.
-0.01 13 years
-0.04 8 years
0.05 2 years
-0.08 6 years
-0.09 28 years
-0.09 2 years
-0.14 10 years
0.15 27 years
0.37 18 years
4 of the 5 that where Zillow was off the most were 10 years or longer since the most recent sale. I’d probably throw out the one they were off the most, because that was the 1.1 million house. But that would still leave 3 of the 4 they were off the most being 10 years or more since a sale.
While the most accurate was 13 years, then next three were an average of 5 years.
Note also only 3 were within 5%, and just over half the Zestimate was lower than the actual sale price.
neither was too bad in this case, but cyberhomes correlates a little bit better (r=0.98) with the sale prices than Zillow (r=0.96).