By Travis Barrington on September 18, 2015.
Originally from the Philadelphia area, Michael Mandel moved to New York right out of college. He had a brief stint working with a commercial music production start-up before becoming a leasing broker at Grubb & Ellis. “I always had an interest in real estate. I like the tangibility of real estate. I like the fact that you can see what you’re working on. I was looking to do something entrepreneurial, and real estate — as far as a job is concerned — is as entrepreneurial as it gets. I enjoyed it but this is more fun.”
Mandel is referring to CompStak, the company he founded with software engineer Vadim Belobrovka in January 2012. Mandel and Belobrovka devised a crowdsourced model in which brokers trade lease “comps” – property data like rent, square footage and tenants – in exchange for other comps.
CompStak’s unique sharing model has been well received by the real estate community who in the past were reluctant to share data with their peers. Before CompStak, data providers would employee dozens of researchers scraping information from unreliable public sources. By using comps as currency, CompStak has found a way to encourage participation. The collective knowledge of the community has also increased the accuracy of the data. “We are all in this together,” commented Mandel.
Today, the company has grown to about 55 employees and has raised just over $10 million in venture funding. They are based in New York but their largest market, data-wise, is the San Francisco Bay Area, followed by LA, and then New York.
This year, CompStak has been focusing-in on revenue, with a redesigned enterprise platform and a new on-demand service. For the first time, users will have the option to purchase lease comps individually for $25 each, without a long-term commitment. The previous model allowed companies to have access to the entire database for annual fees averaging about $35,000. Dozens of deep-pockted firms like Tishman Speyer, Wells Fargo and SL Green subscribe under that model. Smaller firms that only need occasional comps had been reluctant to pay the sizable fee.
The new platform allows users to search an address and immediately receive an in-depth picture of its competitive set. “Any commercial real estate investor, lender, or asset manager can quickly understand how markets, submarkets, and individual properties are performing and then compare that to similar properties,” Mandell said. “We can do that based off of starting rent, effective rent, concessions. And it’s pretty darn accurate.”
CompStak continues to reserve a lot of functionality for their enterprise customers. Not all of the filters or maps are available for on-demand users. For example, the ability to export to Excel is only available to enterprise customers. The ability to edit your comp-set and remove or add certain buildings is also limited to enterprise customers.
Mandel noted that CompStak is migrating from a data tool to a decision making tool with additional insights and analytics that allow users to reach conclusions from data quickly. “We recently released CompStak insights and that goes hand-in-hand with on-demand. It allows you to get insight into any sort of search you’ve done. For example, if you search for buildings on a map that have had deals done in the past year above the tenth floor, we can give you full charts and graphs showing how those buildings performed over time so you can see the trends. It’s an instant market report.”
In addition to their on-demand service and analytics, CompStak is actively working on integrations with other companies — both new and established, as well as preparing for international growth. “We have gathered enough data to launch a London platform today but there are some technical hurdles we still need to overcome,” confirmed Mandel.
Last year, the company surprised many of their clients and partners with CompStak branded M&Ms for Halloween. Their latest novelty may prove to be popular in drizzly London. “We’re sending out CompStak ponchos — because it’s raining comps,” Mandel quipped.