Data Wars

In the last week of October, commercial real estate brokers all over the country received a package via snail mail from CoStar Group. Inside was the informational equivalent of a dirty bomb — the latest salvo in the data giant’s legal battle against its main rival, Missouri-based Xceligent. When brokers opened the UPS envelope, they found a brochure with a carefully assembled selection of news clippings and letters about Xceligent’s alleged theft of CoStar’s data. It also included a congressional letter and news reports about human and sex trafficking.

The bizarre root of that part of the story: CoStar won a court order in 2016 that led to the raid of an Xceligent contractor’s offices in the Philippines. The data giant later claimed to have found evidence indicating that the contractor, Avion, helped facilitate sex trade on behalf of another client, Backpage.

Soon after brokers received the package from CoStar, Alison Melton, an Indianapolis-based retail director at Colliers International, had reached a breaking point.

For days, she had been on the receiving end of a barrage of emails from CoStar and Xceligent — which both maintain large databases of commercial properties, players and deals — with the competing firms making accusations, denials and counter-accusations against each other.

“Dear CoStar Real Estate Manager and @Xceligent: With all due respect, I really don’t care,” Melton wrote in a LinkedIn post. “Please stop sending me your tattling emails.”

Her post, which racked up 150-plus likes, hit a nerve.

“All brokers/agents need to band together, no matter what size company and take our data back over,” one commenter wrote. “What a waste of money, how about lowering the exorbitant monthly subscription costs instead,” wrote another. Melton declined to comment for this story.

In many ways, commercial real estate brokers, landlords and other professionals are like the children caught in the middle of a messy custody fight.

In December 2016, CoStar sued Xceligent for copyright infringement, alleging it stole property images from CoStar databases. In June, Xceligent hit back, filing an antitrust lawsuit seeking to end CoStar’s alleged “decades-long monopoly.”

The legal spat between the two competitors, however, is a symptom of a more profound shift in the real estate industry: the rise of data as a sacred resource.

For as long as real estate firms have been around, they’ve measured their size and importance in square feet. But over the past few years, that’s started to change. Increasingly, they’re also measuring their worth in megabytes.

The war over real estate intel has not only spawned a slew of startups looking to take on Goliaths like CoStar and Xceligent and carve out their own niches; it’s also lured in a growing number of big global brokerages and market unicorns like WeWork. In WeWork’s case, a vault of user and building intel has helped justify a $20 billion valuation.

Perhaps most notably, the real estate data realm has attracted money from major investors including the Blackstone Group, Brookfield Properties and the Durst Organization — all of which clearly believe there are returns to be had.

(Click to enlarge)

“Anyone who is building data or software capabilities in the industry is trying to be a driver in the middle of that ecosystem,” said Bob Courteau, CEO of the Toronto-based real estate advisory, software and data solutions firm Altus Group, which has worked with Ivanhoe Cambridge, JLL and Colliers, among other clients.

While industries like Wall Street had their data revolutions more than a decade ago, real estate is just starting to catch up.

“Compared to other financial markets, we’re still in relative infancy in the use of data,” said Bob White, founder of the research firm Real Capital Analytics, which tracks global investment sales and real estate debt deals.

On the residential side, the fight is over the digital listings market, with giants like Zillow now going head to head with the brokerage world to milk the market for all it’s worth. And though sources say the commercial real estate sector can only benefit from having more information at its fingertips, they predict more legal conflicts and express concerns about laying the groundwork for monopolies.

The rising value in property intel means “there’s increasing risks associated with that data,” said Mike Sroka, co-founder of Dealpath, a commercial real estate deal management software startup. He argued that there will likely be more litigation over the control of data in the years to come. “This is a lot of new territory, which means that the market hasn’t really ironed out all of these things yet,” Sroka explained.

There is also a fear that the brokerages, landlords and other real estate players that are too slow to embrace the trend will pay. The race for data will only heat up, said Chris Kelly, co-founder of the office services provider Convene, which counts Durst and Brookfield among its investors. The New York-based firm has raised $119.2 million in venture funding since launching in 2009.

“If you’re not a company that’s going after data,” Kelly said, “you’re going to fall behind.”

Bits are the new bricks

In some cases, data has become almost as important as the physical property.

Years ago, if a commercial broker needed information about a building, they would find out from another agent or go to the property. But as big, multinational institutions expanded their market share and began investing nationally and internationally, that became less feasible. In addition, with shareholders and investors to report to, those megaplayers needed more numbers and data to justify their real estate investments. Suddenly, data processing became all the more valuable and sexy.

To be sure, real estate firms have always relied on data. It’s just that they long kept records in filing cabinets and later in Excel spreadsheets. Over the past decade, though, the game has changed at a rapidly growing rate. Where New York real estate players once had to trek Downtown to the city’s Department of Finance or the Department of Buildings to get records, for example, that information is now online. And startups have seized on it to create algorithms to churn through permits, tax records and other information so that they can build — and monetize — their own databases.

Richard Sarkis, co-founder of the online property database Reonomy, said that even before CoStar launched in the late 1980s, having so-called “asymmetric information,” meaning intel that a competitor didn’t, was an advantage.

“Now,” he said, “it’s come back almost full circle.”

“You want to buy the best data that’s available out there,” Sarkis said, noting that “there’s real FOMO” — the popular acronym for “fear of missing out.”

Firms such as Honest Buildings and VTS now allow real estate companies to upload their own data, including rent rolls and construction bills, to online platforms to track their portfolios.

As the demand for data and online tools continues to skyrocket, sources said big real estate companies have become much more sophisticated in their use of intel over the past three to four years and are increasingly willing to pour money into digital technology.

Hans Nordby, a managing director at CoStar, said he has noticed more commercial real estate companies talking about how they can use their own data to get a competitive edge. “There’s a lot of talk about information making powerful businesses. I think that’s definitely true,” he said. Nordby declined to talk about the legal tangle with Xceligent, and CoStar would not comment on specific allegations against the company.

Benjamin Berookhim, an executive at JPMorgan Asset Management, said he has noticed an increased focus on data across real estate investment. “The clients are asking about it more, and it needs to be addressed,” he said.

Sources said the global commercial real estate brokerage CBRE had a watershed moment in 2016 when it hired Capital One senior vice president Chandra Dhandapani as its chief digital and technology officer. Bringing on a high-level bank executive to oversee technology and data signified the field’s growing importance in the brokerage business.

“We are focusing on improving visualization of existing data so that clients and internal users can more easily draw conclusions and insights from the data,” Chandra told TRD by email. “We are starting to use machine learning algorithms to provide contextually relevant information.” She declined to go into detail on how CBRE, which has 75,000 employees, plans to monetize those efforts.

JLL has also been investing heavily in the sector over the past few years. The commercial brokerage now has data teams on every continent, said Sean Coghlan, the firm’s head of investor research. JLL, which has close to 80,000 employees, uses that intel to support its brokers and advisers in addition to marketing the firm’s prowess to clients and making it a go-to source for the media.

“Most people in the industry acknowledge it’s a priority,” he said. “A lot of companies are playing catch-up to peer industries that are much farther ahead.”

Representatives for Cushman & Wakefield — another leading data provider among the global commercial brokerages — declined to comment for this story.

Berookhim said JPMorgan Asset Management began its data initiative about five years ago. The financial services subsidiary evaluated what kind of intel it needed and how it could better gather it, then hired an in-house research team and built its own proprietary data-processing platform, Berookhim said. He argued that the asset manager’s size and the scope of its investments allow it to collect useful data such as rent rolls and other financial stats. “The amount of information we have helps us extrapolate trends better,” he said.

Smaller firms have also jumped on the data bandwagon.

Quantierra, a self-described “quantitative” real estate advisory firm co-founded by former RCA executive Ben Thypin, is trying to muscle in with novel ways to use public property intel to generate leads. The firm launched nearly three years ago and has closed on more than $20 million in deals to date.

“Everyone knows the Bronx development market is hot, so we asked ourselves, how can we get in on this action but not run into competition from all the other brokers?” he said.

The firm searched tax records to create a database of which property owners were likely in financial distress, and then contacted them offering help selling properties.

“We’re in contract on two of these sites as we speak and have never set foot on either site,” Thypin added.

The distorted picture

The proliferation of data and data platforms in commercial real estate has created an environment ripe for legal conflicts and a potential golden age for copyright lawyers.

A decade or two ago, ownership of information was clear-cut: If a filing cabinet or computer stood in your office, the data inside was probably yours. But that is no longer the case.

RE BackOffice, a small Pittsburgh-based commercial real estate software and services company founded in 2006, recently learned this the hard way. In October, it suddenly found itself at the brink of annihilation. Since about 2012, Xceligent had been paying the company to scrape the web for property images and information to put in its online database. RE BackOffice thought little of it, since the images were mostly pulled from brokerage websites accessible to anyone, according to a source close the company who spoke on condition of anonymity.

But that’s not how CoStar saw it, according to a federal lawsuit it filed against RE BackOffice on Oct. 19. CoStar accused the company of acting as an accomplice to Xceligent’s alleged “mass piracy.” It alleged that for years Xceligent directed the software provider to trawl CoStar’s databases and steal its images.

Faced with the prospect of a legal battle it could ill afford against a $10 billion corporation, RE BackOffice’s CEO, Harbinder Khera, surrendered. A day after the lawsuit was filed, he admitted to the wrongdoing in a court filing, and CoStar immediately trumpeted the news to its customers.

It wasn’t the first time the data giant seized on the murky ownership of intellectual property in commercial real estate.

“CoStar is the villain here,” said Thypin. “To them, litigation is a growth and market-share strategy.”

Khera could not be reached for comment.

Founded in 1987 by Andrew Florance, the son of a well-known architect in Washington, D.C., CoStar has grown into the country’s most dominant online database for commercial property and leasing information, with a market cap of nearly $11 billion as of Nov. 29. The company has 3,775 employees in 67 offices, and many of its researchers spend their days calling brokers and landlords to collect data, according to sources familiar with the firm.

By all accounts, CoStar has helped make the real estate industry more efficient and transparent, providing a vast amount of information to anyone willing to pay its hefty subscription fees, which can total thousands of dollars per year and person, depending on the product.

But CoStar also has a reputation for suing its competitors.

In 2014, the company filed a so-called John Doe lawsuit against several users of the crowdsourced lease and sales comp site CompStak, alleging they had uploaded and shared CoStar-owned property images without permission. In response to a court order, CompStak handed over their names, and the suit was dropped.

In 2015, CoStar filed a similar suit against RealMassive, which runs an online commercial real estate database. And it sued Xceligent for copyright infringement mere months after that company — which is now owned by the Daily Mail and General Trust — announced a push into CoStar’s biggest market: New York City.

The data giant claimed that photos carrying a CoStar watermark had appeared in Xceligent’s database. Xceligent argued that it never deliberately stole images, contending that CoStar watermarked images actually belonging to brokers and landlords, who shared those same images with the data provider.

Xceligent argued that preventing those players from sharing their images with other firms was anticompetitive.

“I think it’s comical, actually. CoStar has put together a brilliant strategy to defend their moat through photos,” said Nick Romito, the CEO of VTS. “[No other] business in the world has figured out a way to lightly put a watermark on a photo and then have people investigate to see ‘Is there a watermark?’ and if so, sue the shit out of them,” he added. “This is fucking crazy.”

Amy Goldsmith, co-chair of Tarter Krinsky & Drogin’s intellectual property group, said the law is clear: If a broker took the images, he or she owns them — unless that broker explicitly assigned ownership to CoStar. But the equation changes if the images in question were taken by the firm’s employees. “I doubt CoStar would have alleged copyright infringement of 9,000 images if it didn’t own the majority,” Goldsmith said.

Jefferson Scher, a partner in the intellectual property group at law firm Carr & Ferrell, said that “it’s hard for people to keep straight who owns the copyright of an image.” In most cases, the people who photograph properties own the rights unless they sign them away. But they may unknowingly sign them away online through so-called click agreements. “That’s sort of the catch, that we click on things without reading,” Scher noted.

The ambiguity of data ownership has also emerged as an issue for the co-working behemoth WeWork. In July 2016, Thinknum, a tech startup that monitors companies’ websites on behalf of investors, scraped WeWork’s online membership information to analyze its member churn. It then published its findings in a blog post, which was picked up by The Real Deal and other news outlets.

Office services provider Convene, which manages common space in One World Trade Center, is making a big data play to create smarter buildings.

As a member of WeWork, Thinknum had access to the numbers that the public did not and considered that information fair game, the firm’s co-founder Justin Zhen told TRD. But WeWork claimed the startup had violated its membership rules, pressured it to remove the post and kicked it out of its office space.

“Just like Watergate, where you can’t break in and steal files, you can’t break into someone else’s server and steal and scrape,” Goldsmith said.

Zhen argued that Thinknum never broke into anywhere, since the data it scraped was “readily accessible” to WeWork’s members. Representatives for WeWork declined to comment for this story.

Meanwhile, firms like VTS need to perform a delicate balancing act. While their business models are based on collecting and sharing data, they need to assure their customers that they won’t divulge sensitive information. When VTS signs contracts with its customers — which include the likes of Boston Properties and Blackstone — it typically reserves the right to aggregate their data to calculate figures such as average market prices without revealing anything about specific companies or buildings.

Romito said that in the company’s early days, many of its customers balked at sharing any data at all. Now, he said, they’re seeing the potential benefits and coming around. Last year, VTS merged with its rival Hightower in a blockbuster deal valued at $300 million that cleared a big potential obstacle to growth.

Eventually, VTS wants to turn its customers’ anonymized data into a research product it can sell, but Romito said that might still be years away. “We want to do it when the market is really ready, and when they’re comfortable,” he said.

RCA’s White said murky ownership rules and the desire among real estate firms to keep data out of the public eye will likely to lead to more legal disputes going forward. “There’s a lot of gray areas in terms of who owns what,” he said.

CompStak co-founder Michael Mandel said he’s more optimistic, drawing an analogy to the music industry, which saw a flurry of lawsuits in the late 1990s and early 2000s as big record companies fought against new online music-sharing platforms like Napster and Limewire. Eventually, as courts clarified copyright rules, the dust settled, allowing services like Spotify and Pandora to exist.

Mandel said he thinks something similar may happen in the real estate data industry. “There’s going to be more rules over what’s acceptable and what’s not,” he said.

Banking on “big data”

Convene’s Kelly has a vision of commercial real estate’s future, which he calls the “empathetic” office building. Once someone enters a room, the building will recognize them through a signal from their cell phone.

Lights, desk height and temperature will instantly adjust to their preferences. Perhaps a message is sent to the concierge to order coffee just the way they like it. While today workers in the same office generally all have the same experience, Kelly said, their experiences will be individually tailored in the future — all thanks to data collection.

“In 10 years, you’ll walk through an office and be like, ‘Oh my God, I remember when I had to turn on every single switch and everybody was either too hot or too cold and everybody had the same desk and had to sit next to the same person every day,’” he said. “And you’re going to laugh about that.”

All of that speaks to the second commercial intel revolution underway: big data. 

While firms like Reonomy and RCA mostly use existing information — building measurements, rent rolls, comparables and debt and equity investments — others like Convene are collecting completely new types of data.

Convene’s main business is managing common amenities such as conference rooms, catering facilities and co-working spaces. But for the last 18 months or so, Kelly’s team has also been quietly working on software that would function as an office building operating system and eventually serve those “empathetic” buildings. The company is beta-testing the software at two undisclosed Brookfield properties, with plans to have it running at 25 office towers by the end of 2018.

The firm is also in the midst of a $150 million Series D funding round and plans to spend a big chunk of that money on the operating system, Kelly said.

Meanwhile, CBRE is making its own big data play, experimenting with beacon technology — a wireless Bluetooth system that’s been used in the retail world to send signals from cell phones to receptors to track customer movement.

And then there’s WeWork, which currently operates a test floor in its Times Square location at 1460 Broadway. The company says the data it collects on its members’ behaviors will give it a crucial competitive edge. And a fawning May profile in Wired magazine put that intel at the center of WeWork’s business model, arguing that the co-working behemoth “is wrangling data that it’s collected over the years to quietly reorganize an office around its workers.”

During a 2016 tour, David Fano, WeWork’s chief product officer (now chief growth officer), told TRD that the company was looking to create heat maps of where people work and congregate — information it could then use to design future office layouts. By tallying the traffic through certain doors, for example, it can determine which spaces are most and least useful and then modify its floor plans, which would presumably make the company more attractive to members and investors.

It could also use the technology to send members alerts for add-ons, which will help increase profits. If a member walked into a conference room that he or she hasn’t booked, for example, WeWork could instantly send a message to their phone offering to rent it.

Not everyone is convinced that gathering and using data on how office users move is really a game changer. “It’s a completely minor advantage,” said Zach Aarons, co-founder of the advisory firm MetaProp NYC, which operates a real estate tech accelerator and seed fund.

The making of monopolies

While the real estate industry has long been fragmented and inefficient, in many ways it’s also been remarkably egalitarian. Smaller brokerages like Eastdil Secured and Rosewood Realty, for example, go head to head with the global giants including CBRE and JLL in major markets.

But some believe the rise of data could turn that meritocracy on its head.

“It will be harder for smaller players to compete going forward,” Kelly said.

By virtue of their sheer size, big firms have more data. And if the largest investment sales firms were to figure out how to use their intel to predict which properties will likely appreciate in value, for example, they could serve their clients far better.

“Why would I ever trust anyone else to help me find property?” Aarons argued. “A small brokerage can’t do that. They don’t have the data and they don’t have the resources to build that kind of technology.”

The uneven playing field has even bigger implications for data and software companies. 

That’s because the bigger a database is, the more valuable it becomes to consumers. That reality virtually ensures that the dominant players like CoStar keep growing. Xceligent, meanwhile, has been losing steam at a rapid clip. In late November, Daily Mail and General Trust announced that it had dropped the money-losing data subsidiary’s carrying value to zero amid “disappointing” revenues from its push into New York.

“Data networks are inherently monopolistic,” Sroka said. The same issue has dogged the social media space, where Facebook quickly became the alpha dog; the e-commerce arena, where Amazon reigns supreme; and the online search world, where Google has virtually no credible rivals.

Scher of Carr & Ferrell argued those companies’ blueprints for success could be replicated by firms in other data-centric industries. “Google knows how people think,” he said. “If you want to compete with them, you can’t compete with the behavioral aspect of that data.”

If a firm were to compile a trove of behavioral data on office usage, it could also gain a strong competitive advantage. “Why not let some company be awesome at doing this?” Scher proposed. “It’s just a question of ‘Can we sustain a capitalist economy if it turns out there’s always only one company that wins every single industry?’”

In spite of conventional wisdom, Sroka argued that monopolies may be a good thing in this context. For many people, life is easier because everyone is on Facebook and because a web browser and search engine are run by the same company. The same principle applies to real estate data, he said.

Many already see CoStar as that commanding player in commercial real estate’s data revolution. “They’re a monopolist, basically,” Thypin said. Leasing brokers may grumble about the data giant’s steep prices, but they depend on its products and often have no viable alternative.

Some argue that there is one crucial difference between CoStar and Google, however. While the search engine giant collects its own data, CoStar depends on real estate companies to provide certain intel.

“It’s as if Windows was built on code provided by its customers,” Thypin said.

As real estate firms increasingly recognize the value of their data, they may become more reluctant to give it to CoStar for free only to have the company effectively sell it back to them in form of a subscription, Reonomy’s Sarkis argued.

“That’s going to create some cognitive dissonance, I assume,” he said.

To Sarkis and others, this means firms have a chance to compete with CoStar if they use novel technologies or grant users more control over their own data.

“Any incumbent is an incumbent for only so long,” Sarkis said.