While automated home value estimates may be useful in certain applications, they can not take the place of the expert pricing advice of a real estate agent.
Remember a few years back when that plane’s autopilot algorithm made a perfect emergency landing on the Hudson River in New York City?
Me neither. That’s because an algorithm didn’t land that plane. An experienced human pilot did. If you’re wondering what this has to do with Lakes Region real estate, bear with me and read on.
On January 15, 2009, US Airways Flight 1549 was rocked by a rapid succession of explosive bangs less than three minutes after takeoff. It was the unmelodious percussion a commercial jet makes while smashing through a large flock of Canadian geese at more than 200 miles per hour.
The plane was about 2,818 feet above New York City when its twin engines instantaneously “ingested” multiple birds roughly the size of Thanksgiving turkeys. The engines belched flame then went dead silent. The cabin flooded with the odor of fuel.
The pilot in command, Chesley “Sully” Sullenberger, issued a Mayday: “This is Cactus 1539, hit birds. We’ve lost thrust on both engines. We’re turning back towards LaGuardia.” But Sully quickly realized they’d never make it back. He considered Teterboro Airport in New Jersey but made another snap decision based on many years of education and experience. He radioed: “We can’t do it … We’re gonna be in the Hudson.” The aircraft began a 240 mile per hour “glide descent” and Sullenberger addressed passengers and crew with three little words you never want to hear from your pilot: “Brace for impact.” Sully then “ditched” (the euphemism for crash landing a powerless plane not designed for water landings) the commercial airliner into the Hudson River at approximately 140 miles per hour.
I’m certain many of the geese involved didn’t make it, but all 155 passengers and crew aboard Flight 1549 survived. The incident became known as the “Miracle on the Hudson.” Stephen Rice, Ph.D., an aeronautics and robotic flight expert with Embry-Riddle Aeronautical University later admitted, “No autopilot today could have landed the plane on the Hudson River.”
So what does this gripping introduction have to do with Lakes Region real estate? In real estate, as with commercial aviation, technology is a useful tool. But there is simply no substitute for human experience and judgement.
Millions of home buyers and sellers have been engrossed with the home value equivalent of autopilot. Just visit a major real estate dotcom like Homes.com, Zillow or Redfin, enter an address and instantly get a free estimate (based on “proprietary algorithms”) of that home’s value.
But air passengers rely on experienced professionals (not just sophisticated technology) to get them where they want to go, and home buyers and sellers should too. Let me tell you why.
Last year an independent research firm published the results of a nationwide study of home sales and reported that the Redfin Estimate “more accurately predicted the value of thousands of homes for sale” than estimates by Zillow and Homes.com. Of the 5,074 homes sales evaluated in the study, 64 percent sold within three percent of the price predicted by Redfin, compared to 29 percent for Zillow and 16 percent for Homes.com.
But if you analyze those numbers critically, this study really shows that Redfin offered the least inaccurate automated prediction of home value.
Redfin’s predictions were three or more percent off closing price 36 percent of the time. Zillow’s Zestimate® was outside that range with 71 percent of its predictions, and Homes.com predictions were more than three percent inaccurate a whopping 84 percent of the time.
So think of it this way… would you feel confident flying an airline with a 36 to 84 percent likelihood of missing your destination airport’s runway by at least three percent? Then how much trust can you place in algorithms scientifically proven to guess home value too high or too low 36 to 84 percent of the time?
The study also reported that the median prediction error rate (how much estimates exceeded or fell below actual sale price) for Redfin was 2.06 percent off actual sale price. Zillow had a 5.95 percent median error rate, and Homes.com’s was 10.26 percent.If you apply those error rates to the median closed price for single family homes sold in New Hampshire in 2017, Redfin’s autopilot guess would be off by $5,479 in home value. Zillow’s would be off by $15,827, and Homes.com would be $27,291 too high or too low. (For reference, as a whole single family homes sold in New Hampshire in 2017 closed 1.46 percent below listed price.)
Don’t get me wrong, technology is a powerful tool. Real estate is a data driven career and REALTORS® depend on digital SLR cameras, drones, smart phones and mobile devices, computers, web applications, satellite imagery, databases, CRM systems and a variety of software. But we ultimately rely on human judgement and experience because we know that your home’s sales price is likely to depend upon many factors alien to algorithms, including the property’s condition, the motivation and means of buyers and sellers, the marketing skills of the agent, and the psychological impact of supply and demand levels.
Automated estimates can be useful, but I believe home valuation is more challenging in New England, where homes built and added onto since the 1700s can neighbor ones constructed in the 21st Century. The quality of land and view here can vary dramatically between the subject property and a comparable home a half mile away. Algorithms that are accurate in metropolitan markets, where builders quickly erect hundreds of standardized condominiums or single family homes based on the same materials and handful of models, can struggle with the unpredictable variables New Hampshire’s rural real estate market is full of.
The bottom line is that technology is a useful tool but in real estate as with airlines there is no substitute for human experience and judgement. That’s why Redfin’s website says its automated Estimate “is just a starting point—it is not an appraisal or a substitute for the expert pricing advice of your real estate agent.”
In other words, if an estimate with a 36 to 84 percent chance of being off the mark is good enough, trust autopilot pricing (did I mention that Lawrence Sperry, the inventor of autopilot, fatally crashed his plane into the English Channel?). If you’d like what Redfin calls “expert pricing advice,” contact me.
Please feel free to visit www.rocherealty.com to learn more about the Lakes Region and its real estate market.