The face of big data

The loosely defined term “big data” is on the lips in real estate offices around the world these days, describing sets of facts and figures that are too large and complex to work with using nonspecialized computers. The industry’s so-called “big data gurus” are designing algorithms parsing the data that do everything from calculate a home’s market value to match commercial spaces with ideal tenants.  

Stan Humphries, chief economist and chief analytics officer for the Zillow Group, is the O.G. who defined the role of data guru. He is the man behind the first Zestimate algorithm and was instrumental in other tools that have attracted millions of users to the platform, helping to make Zillow the $10 billion company it is today. 

Humphries set a precedent for the role of chief economist in the real estate space — someone who could crunch numbers, spot trends and then hop on a call with a reporter and explain it to the average American, thereby driving Zillow brand awareness sky-high. 

“He certainly set an industry standard for what it means to be a chief economist,” said Ralph McLaughlin, former chief economist for Trulia, which is owned by the Zillow Group. “Some of his work is the most respected out there, and he set a standard in terms of his ability from an external perspective to be a very clear and coherent expert in the housing market, and that’s not easy.”

The key to getting the public to appreciate and understand the products and the insights Zillow produced was Humphries’ ability to put a cogent message together, sources said.

“Stan was certainly a pioneer in the space because he had this mixture of solid data analytics, this ability to take things from economics and forecasting to figure out what this means for the market, and also an element of storytelling,” said former Realtor.com chief economist Jonathan Smoke.

A zap to the market

There’s no greater example of the tectonic shift big data is capable of than Humphries’ Zestimate, an algorithmic home pricing feature launched 12 years ago, when “big data” was a term rarely heard beyond the campus lawns of Silicon Valley.

The Zestimate is calculated using millions of points of publicly available and privately submitted data, like square footage, sales comps and property tax data. It awakened the residential real estate industry to the possibilities of big data and armed potential buyers with a wealth of information — much to the chagrin of some brokers who have long argued that the Zestimate can be wildly off-target, projecting values that can be over or under a home’s actual market value by hundreds of thousands of dollars.

But like it or not, the Zestimate had a huge impact on the market. “It’s a topic of conversation in nine out of 10 meetings,” said Los Angeles-area Pacific Union International agent Sonia Cabrera, who said she saw an immediate impact on the L.A. market when the Zestimate came out in 2006. “Before, it was based on postcards and phone calls from agents.”

The man behind the formula had zero experience in real estate prior to Zillow — something Humphries admitted to Spencer Rascoff when the CEO first asked him to come aboard to “figure out what every house in the country was worth” in 2005.

At that time, Humphries had fairly recently gotten his Ph.D. from the University of Virginia’s government program and was running analytics at the travel website Expedia, where he met Rascoff. The CEO said in an interview last year that he shrugged off Humphries’ concerns about knowing nothing about real estate. “This is a math problem,” Rascoff said he told Humphries. “It has nothing to do with real estate.”

At Expedia, Humphries was creating models to better decide what products to show customers, as well as forecasting supply and demand at hotels. He was also interested in public policy, earning a master’s degree in foreign service at Georgetown University and working as a presidential fellow for the science adviser in the Clinton administration. He showed an early interest in using data to inform public policy, which proved crucial for Zillow a few years down the road.

“He wanted to get into big data before it was popular; he was one of the innovators,” said Humphries’ Ph.D. supervisor, Steve Finkel, who is department chair of political science at the University of Pittsburgh.

After launching Zestimate, Humphries continued to develop algorithms and frequently shared market reports and analyses on Zillow’s blog, Porchlight. The algorithms he helped build allowed him to quickly tailor data to spotlight interesting trends in local markets around the country — something that only the largest national brokerages could do at the time. He frequently highlighted Zillow’s products in those analyses, like the Zillow Home Value Index, the calculation of which is a mystery to those outside the company. 

Media in local markets across the country often picked up reports targeted to their regions, and Humphries made himself available to them, which helped spread Zillow’s name and attract users, said his friend Susan Athey, an economics of technology professor at Stanford Graduate School of Business. 

“It was impressive that [Humphries] was able to build models and also talk to the media,” Athey said. “A standard data scientist might have a harder time talking to an economics reporter at a newspaper, but people like Stan are able to also bridge that gap.”

Athey teaches the mechanics behind online marketplaces and online advertising. As part of the first wave of tech economists on the scene during the internet explosion at the turn of the century, she has advised companies including Microsoft, where she is the consulting chief economist today.

Historically, economists had mostly been concerned with the macroeconomics of the business, the so-called “big picture,” and provided useful forecasts and analysis to decision-makers, but Athey and other economists at Silicon Valley pioneers like Google began carving out new paths in the field.

“In the early 2000s, starting with Hal [Varian, chief economist] at Google and myself at Microsoft, we saw the growth of a new type of chief economist who focused more on microeconomics issues — data-driven business decision-making as well as marketplace issues, marketplace design and policy,” she said. That’s exactly the position Humphries created at Zillow, which from his example eventually became popular industrywide.

The company kicked its public outreach up a notch in 2008, and that’s when Humphries’ star really began rising. Amid the housing crisis, leadership at Zillow saw what Rascoff later called in a Washington Post interview the “massive void in academia, government and in the media” for “unbiased data” in real estate. He said that data from trade associations like the National Association of Realtors failed to reflect the reality of the declining market.

By making the data held dear by real estate professionals as transparent as possible — with the notable exception of sharing the proprietary Zestimate algorithm — Zillow aimed to captivate U.S. homebuyers and sellers. And of course, the more people who logged in curious about their Zestimate, the more eyes on the ads Zillow sold.

Humphries was promoted from vice president of data and analytics to chief economist in 2009. He and Zillow went to Washington, D.C., and presented data from tools like the Zillow Home Value Index to public policy groups and lawmakers, hoping to be their new go-to data source.

Establishing a presence in the capital was a canny move. In 2011, Humphries was called to testify before a congressional committee about ways to address the 1.5 million homes in foreclosure at the time. Likewise, in 2012, Rascoff moderated a Google Hangout with then-Secretary of Housing and Urban Development Shaun Donovan and a group of homeowners.

Zillow’s public profile hit a peak in 2013 when Rascoff secured a one-on-one interview with then-President Barack Obama about the latter’s housing policy plans. The announcement of the interview caused Zillow’s stock value to spike by around $7.29 per share, or $250 million total, which brought the valuation of the company’s 34.5 million outstanding shares to upwards of $3.3 billion.

“It was absolutely part of our marketing plan,” Rascoff said in the Post interview of Zillow’s Washington outreach. “As a result of being the biggest real estate site on the Web, increasingly politicians are coming to us to say, ‘Hey, how can we get the word out, and how can we interact directly with homeowners?’”

Resurgence of an old-school role

Eyeing Zillow’s wall-to-wall media coverage, competitors in the last several years have been adding chief economists to their own rosters in the hopes of mirroring Zillow’s success.

News Corporation’s Realtor.com, a listings platform with close ties to frequent Zillow critics at the National Association of Realtors, hired Smoke as its first chief economist in July 2014.

Each chief economist at various entities could offer something unique to the market thanks to access to their company’s data sets, Smoke said. For example, he could provide more insight to the supply side of the market, because Realtor.com had access to “practically all of the inventory in the country,” Smoke said.

“We were all competing with each other in that arms race,” Smoke, who left the company in 2017, told TRD. “We were all very firmly encouraged by our organization to be thought leaders, producing insights that were relevant,” he said.

But not everyone is 100 percent sold on the need for a data guru.

Rich Sarkis, CEO of Reonomy, which provides data tools for the commercial real estate industry, has a big tech and data team but no public-facing chief economist. The company recently closed a Series C round of fundraising, so he and his co-founders are thinking of bringing someone in, but it’s not a sure thing.

“There’s a different school [of thought] that says, ‘Hey, if you’ve got the data and you make it so easy to access, you empower the user to become the economist themselves,’” he said. “Then they can start to slice and dice your data. The end user becomes the evangelist.”

Investors hungry for proptech

The opportunities presented by harnessing big data in just the right ways have in recent years opened up a new revenue stream for real estate tech firms, but whether they’ll actually turn the efforts into profits largely remains to be seen. 

VCs sank $18.6 billion into real estate property technology — known as proptech — firms from 2015 to 2017, and more than 25 percent of that came in the fourth quarter of last year alone, according to a report from real estate tech research and marketing agency Re:Tech.

The Japanese firm Softbank poured $4.4 billion into WeWork last year, and Compass — which recently announced that it will begin licensing its technology to other firms — has raked in $550 million in funding ($450 million of which also came from Softbank). 

Commercial brokerages like JLL and Colliers International are also getting into the game, launching their own proptech accelerator labs in which engineers design specialized software products fueled by their respective mountains of market information. 

Sarkis said the idea is to create a product that’s “very heavyweight on the back end but intuitive, elegant and easy to use on the front end.”

“[Investors are] really looking for something that can bring the data monster to heel,” he said. 

Correction: an earlier version of this story said that Zillow’s stock value rose to $250 million following Spencer Rascoff’s interview with President Barack Obama. It went up by $7.29 per share, or $250 million in total, to around $3.3 billion.