The truthiness of Web 2.0

Zillow is getting a lot of buzz. The site, which provides free estimates of home values, got big writeups yesterday at Cnet and the New York Times as well as a ton of echo coverage in the blogosphere. The company has also scored $32 million in venture financing.

I love the idea of Zillow. Opening up the supply of data about real estate puts a dent in the Realtor cartel, and anything that weakens that price-fixing monopoly is to be welcomed. Zillow’s site itself looks good, too. It assembles all sorts of information on house specs and prices as well as various graphing, mapping, and analysis tools. It is, as Paul Kedrosky writes, “data porn for middle-aged people.”

There’s just one problem. The data’s not reliable. As soon as I got onto the site, I did what I’m sure millions of other people did yesterday: I zillowed my own home. The house description, alas, was way out of date. It didn’t include an addition that we put on three years ago, which meant that the room count, bath count, and square footage were all off. There were also some weird errors. Zillow said, for instance, that the house’s “exterior material” was stucco. I live in Massachusetts, which is not exactly the stucco capital of the world. My house has been clad in shingles since it was built back in the fifties.

If you do some digging at the Zillow site, you’ll find a lot of disclaimers, like this one in all-caps: “ZILLOW.COM PROVIDES THE SERVICES ‘AS IS,’ ‘WITH ALL FAULTS’ AND ‘AS AVAILABLE,’ AND THE ENTIRE RISK AS TO SATISFACTORY QUALITY, PERFORMANCE, ACCURACY, AND EFFORT IS WITH YOU.” There’s also this warning about the quality of the data: “We get this information from the tax assessor in your county so their records may not be up-to-date. To change the details about a home, or the tax value or assessment figure, you’ll need to contact your county assessor’s office.” (To Zillow’s credit, it gives you a personal worksheet where you can tweak the data on your own home, but that’s cumbersome, and it doesn’t help if you’re trying to assess other homes – or if someone else is trying to assess yours.)

The flaws in the data don’t stem from Zillow, but from Zillows’ sources. But users don’t care who’s at fault. They just want accurate, dependable information, particularly if they’re doing something as important and nerve-wracking as buying or selling a house. And the fact that Zillow is just aggregating the data means that it has no way to check or fix the data. The value of its service is at the mercy of its sources.

Zillow’s flaw exposes a deeper problem with the Web 2.0 mashup model. Entrepreneurs are launching all sorts of sites and services that are built on data that they’re siphoning out of third-party sites and databases. Sometimes, the secondhand data is good; sometimes, it’s not. The process of chopping up and bundling data from many different sources can, moreover, amplify inaccuracies. Combine bad data from two different sources and you may get bad data squared. Unfortunately, to the user, the inaccuracies are invisible. The slickness of the interfaces of sites like Zillow, with their tables and graphs and calculators, gives the appearance of credibility, even if it’s not warranted. To use the memorable phrase coined by comedian Stephen Colbert, Web 2.0 sites can all too easily supply “truthiness” rather than truth.

There’s one line of thinking about the unreliability of web information that says, essentially, “get used to it.” People are just going to have to become more sophisticated consumers of information. That’s nice in theory, but it doesn’t wash in the real world. It’s like selling wormy apples and telling customers that they’re just going to have to become more sophisticated eaters of apples. Fruit buyers don’t like worms, and information seekers don’t like bad facts.

The danger is that if users keep coming up against unreliable data they could end up rebelling against the entire mashup model. In Zillow’s case, that would be particularly sad. If the Realtor cartel can use the errors in mashup sites like Zillow to promulgate the message that “you can’t rely on the internet,” such sites could end up reinforcing the status quo rather than breaking it. Truthiness hurts.

15 thoughts on “The truthiness of Web 2.0

  1. Greg Yardley

    Zillow doesn’t want to compete with real estate brokers – it wants them to buy advertising. Therefore the results need to be accurate enough to motivate people to act but not so accurate that they eliminate the need for an accurate appraisal.

    I’m not saying Zillow *could* make their results more accurate – it’s hard to detect mold or a cracked foundation through an algorithm – but if they could, would it be in their interests?

  2. JohnO

    If Zillow gets critical mass, it could eventually take over Realtor aggregator programs (which are horrible and old I’m told). Then it could have an alternate source, the realtors, inputting data into the system – and that data would be much more reliable as it is in the best interest of the realtor to get it right. However, getting realtors to open up their data to the world that now doesn’t need there service, they need some motive and reward. Hopefully Zillow can make it worth the information gatherer’s time to get it right – whether the gatherer is the government (which I’m sure you’re all laughing at) or private business.

  3. Kim Greenlee

    Not only was the data wrong about my home, the map point was off by a couple houses as well. A few years ago I did some consulting for Property Data Systems, a real estate research company in Atlanta, GA. Their main product was called MapMerge and basically did what Zillow does but using a Windows client. PDS got their data and maps from the counties. A lot of information about a property is in the public domain. And as you state, usually out of date and not that accurate. PDS regularly made modifications to the data because it wasn’t accurate.

    The Zillow map looks a lot like a Google map. I don’t know where they get the map image but I do know that Google doesn’t have any problems finding my house. There is no doubt that not associating the data with the actual house is a bug that must be addressed. My concern for the company is about what their options are. PDS only supported the realtors in the counties around Atlanta and a few in Alabama because cleansing and normalizing the data, and creating the map points was so labor intensive. If Zillow is lucky people will make the data accurate for them, if not they will have to do it themselves and that will be expensive.

  4. Jim Walker

    The price for our house was off by quite a bit, but the customization feature actually got us pretty close to our current asking price.

    Their real killer app IMO is the stock-like graphing function that automatically plots the price of our house in terms of two key dimensions – price over time – and price relative to all the houses in our township, city, and state.

    Next time you read about the housing bubble bursting, just go to Zillow and see for yourself – across any region in the country.

  5. Randall Wilson

    The zillow price of our home was about two million too high. I live in an older, but nice, neighborhood in Utah. Most of the homes around us that have sold have gone in the 300K range. Well according to zillow, it is time for me to cash out in a big way.

    -Randy

    http://www.4mysales.com

  6. Filip Verhaeghe

    Isn’t one of the key benefits of the Web 2.0 user intervention? If Zillow were to store the customizations as corrected data for everyone to see, wouldn’t that solve their data correctness problem?

  7. Dion Hinchcliffe

    Nicholas:

    Yes, I’m a little surprised to find myself agreeing with you wholeheartedly.

    Fortunately, there are things like transparency and traceability that might ameliorate this. If not, mashups will be consigned to the also-ran bin of the Web; people just won’t trust them.

    More details that I wrote up here on ZDNet.

    Best,

    Dion

  8. panlibus

    ‘Truthiness’, mashups, and a different perspective

    Dion Hinchcliffe offers a thoughtful response to a post by Nicholas Carr that I almost responded to myself this morning. Dion does it better. One thing I would add, though, is that opening up our closed silos of data…

  9. tecosystems

    Zillow: My Favorite New Software-as-a-Service Application

    If there’s a better way to introduce me to a new software as a service offering than by informing me that my loft is worth a hundred grand more than what I paid for it last spring, I don’t know…

  10. Timothy

    I don’t know how much of this is “truthiness,” but I did run three houses through the system–my parents in Northern Virginia, a recent sold house that I’d renovated in Colorado and a home I just purchased in the Boston area and the output, after adjustments to the room configurations and such, was fairly (

  11. Jonathan

    For whatever a “me too” is worth…

    Zillow had everything wrong about our home, and in ways that would be seriously detrimental to any efforts we might make to sell our home. I wonder if there’s any legal recourse.

  12. Tim

    Jonathan,

    Zillow clearly states (and so does the article) that their data is from publicly available tax assessor sources.

    Here’s the catch — if you update the local tax assessor’s records, your house will appear correct in Zillow. But the local tax assessor will also hit you with a bigger tax bill b/c they know now that your house has been upgraded and is worth more.

    I wouldn’t have a problem if Zillow let owners (again, an authentication problem here) modify their house data) as long as they said buyers would have to verify the data themselves.

  13. Brad Mattison

    I’m a little unclear on why you are linking this truthiness to Web 2.0. I think it might be slightly unfair to give Web 2.0 a bad reputation on the basis of a single website. Web 2.0 actually probably pertains less to data than it does to media and the way in which it is presented. I do understand the argument you make that the successful design and infrastructure of Web 2.0 can be abused in such a way that it is blinding users from the missing and faulty data. But so far, I have only seen this one instance; even though this is a very interesting and useful site.

    A far more interesting thing–I was disappointed to see that Zillow did not support Safari, which I, as a web-programmer/designer, am shocked about. I find Mozilla one of the hardest browsers to comply with, and hopefully, with the Web 2.0 “revolution” (if that’s what we shall call it), more compatibility and standards will replace the inconsistancies we face today.

    Very interesting concept, I was just somewhat disappointed by your negativity towards Web 2.0. I just hope that constructive criticism like this will raise the standards of the internet.

  14. CJ

    We recently purchased a charming historic home in Hammond Heights Historical District, on the west side of Worcester, Massachusetts. We love the old world European flavor of neighborhood of over-sized Queen Annes and Painted Ladies (old pale pink Victorians!) ..such a charming gorgeous neighborhood.:) However! I searched our home on Zillow too and much of the data was completely inaccurate! (I emailed Zillow with a how-dare-you type flavor LOL…they told me to contact our assessor’s office, also)). For example, our home (complete with our c. 1908 marker which we love!) has 2 main stories, Zillow says 1! We have 4 bedrooms, Zillow says 3. We have cedar shingles, Zillow says stucco. Zillow should not display incorrect data! Now, Zillow tells people they can update their own data, using your credit card to prove you are the owner! Who wants to put their CC on the Zillow site! Not us. Zillow: Delete yourself!

Comments are closed.