It’s a number that the homeowner, buyer, and seller are waiting for, and a number sometimes declared with near-reverential certainty by the housing market. If you’ve just surfed your way to Zillow.com to check an assessment of your house or a house you may wish to buy or sell, you’ll see what I mean. The Zestimate, the number that Zillow so prominently displays for virtually every house in the United States, is claimed to tell you exactly how much that house is worth right now. That number, plucked from thin air – the air of the internet, at that – gets cited in local newspapers and on other websites. It’s quoted as if it’s gospel in discussions on national television shows and it’s even sometimes used as evidence of housing-bubble activity in scholarly papers. People love that number. Or at least they love to hate it. It makes them happy, sad, angry and, maybe most importantly, confused. As many realtors and homebuyers can tell you, it is often way off target. Zestimates are more like, well, a mirage: something that looks real, that you can see, that you can almost touch. And maybe that’s because Zillow does indeed monitor millions of ‘facts’ regarding America’s housing stock, yet those facts are drawn from private transactions, commodity transactions that even companies like Zillow never see. Or at least they never see the whole story. In the continually evolving dance of the housing market, where perceptions drive decisions that drive the market, Zillow has inserted a significant new element to that dance. Whether the dance of house-buying, selling and valuations will continue on the Xbox after a move down the block might still remain a mystery. But the Zestimate suggests that, as long as people are willing to believe it, evading detection will be more difficult than ever.
By 2006, Zillow had first emerged and, after that, everyone interested (or in the least curious) in real estate was going to it religiously. Zillow had an attractive angle: it promised to pull back the curtain. For the first time, the average Joe would have unfiltered and boundless amounts of data at his beck and call. No longer would a buyer or seller have to go through real estate agents and appraisers whose specific processes and methods may seem opaque and questionable. Through the might of a few point-and-clicks, anyone could know how much any given home is worth any given second. At Zillow’s unveiling, it offered a shining new diamond mine of data in the form of the Zestimate: a magic single number calculated by ‘secret sauce’.
What makes the Zestimate compelling is that it’s so simple. In a market where prices can vary from one side of the street to the other, the Zestimate is a clean, easy-to-understand number. It’s the promise of reducing the process of buying or selling a home to one number. And that’s what makes it a problem. The market’s not simple, and reducing it to one number is a disservice, at best. At worst, it’s dangerous.
Ponder the mid-century bungalow in an area rapidly gentrifying. Sure, the Zestimate could look at recent sales in the area, the importance of the parcel, the number of bedrooms, and a whole lot of other quantifiable factors. But it can’t possibly encompass the fact that the house has been lovingly redone by a local architect with a trained eye for period detail. Nor can it ever get at the otherwise ineffable appeal of the neighbourhood – not only at the hipness of the mix of current long-timers and newcomers, their warmth and open-heartedness, the sheer fizz of the place that defies reflection in property records. The result could be that the Zestimate also reflects what one of our neighbours called ‘HGTV valuation’, and is way off: below what the market will bear.
By contrast, a Zestimate is likely to provide an overvaluation based upon stale data or just a swath of raw material that the algorithm works with. Maybe it doesn’t yet know that the formerly high-performing school district is slipping from grace, or that a new, cloistering freeway extension has turned the neighbourhood into a drone-noise corridor. In that case, the Zestimate tacks a bill that inflates the underlying value of the home, which in turn leads the owner towards an inappropriate high-watermark for asking price. It then hangs on the market much longer than is reasonable.
The Zestimate’s flaws notwithstanding, it influences the housing market tremendously. For most would-be buyers and sellers, the Zestimate is an opening gambit. A homeowner looking to sell might use it to support a lofty asking price, while someone shopping for a home might hold it up as a basis for lowballing a bid. Real estate agents, who often balk at the Zestimate’s shortcomings, are the first to admit that they have to defer to its wide influence. It.
Just one Zestimate doesn’t move the market; but if Zestimates in an entire neighbourhood or even city skyrocket, market momentum can multiply the effect of the Zestimate on individual transactions. As Zestimates go up, those who are afraid of missing out might jump into the market to avoid getting priced out. Meanwhile, if real estate-obsessed homeowners persuade each other to sell at what they now see as a frenzied peak, it can cause more selling at even higher prices, further boosting the Zestimate. Conversely, a Zestimate fall can cause a wave of panic-selling and still more downward pressure on prices. For those in financial trouble, a foreclosure can become a potentially better deal than holding onto an underwater mortgage. It may be as if the Zestimate not only reflects but actually drives the movement it attempts to merely mirror.
Zillow, for its part, knows the limitations of the Zestimate. Just about every line of text on the site’s ‘About Zestimate’ page warns the reader that the Zestimate is just that, a Zestimate, and that one should look upon the valuation ‘as a starting point in determining a home’s value’. The site has tinkered with the algorithm in dozens of ways since it developed it in 2005, often giving realtors a mild heart attack, trying to improve its precision every step of the way. But whatever their sophistication, no algorithm will ever be able to embed itself in the complexities of the real estate world. The Zestimate is based on data. Data, by its nature, is always incomplete. It can tell you square footage, but not quality of workmanship. It can tell you what comparable homes sold for, but not why. Was it a bidding war, a fire sale, a bit of a feud in the family?
And yet, the Zestimate stubbornly persists, in part because it lends people something to grab onto amid all the instability. The housing market is extremely fickle, and the Zestimate is at least constant, an estimate that, if inaccurate, is at least regular. It’s a starting point in an ocean of unknowable stuff, a reference point for homeowners, buyers and sellers to anchor themselves amid all the competing fictions.
Furthermore, the Zestimate plays to our underlying need for certainty and control. In an arena where almost everything is out of our hands – interest rates, macroeconomics, the moods of prospective buyers/sellers – the Zestimate gives us power. We get to feel we’re in command, that we can quantify and measure the things we own, in even if those things prove impossible to measure.
Ultimately, though, Zestimate might not so much reflect what is happening in the housing market as expose the desires and fears we bring to the marketplace. The Zestimate, then, is little more than a tool: useful in some contexts, misleading in others. It is also a useful baseline in what is an otherwise murky, opaque process. Which is why, for all its wrongness, the Zestimate is worth following. Even if it doesn’t tell us the whole truth, it tells us enough to sustain our interest, to keep us digging and keep us dreaming of the perfect home at the perfect price.
