Real Estate’s New Mantra: Using AI to Predict Property Value (and Returns)

By Pragati Verma, Contributor

When Israeli startup Skyline AI and its undisclosed investment partners heard about an opportunity to buy two multifamily residential complexes in Philadelphia earlier this year, they turned to their proprietary artificial intelligence-based predictive analytics platform to assess the property’s value.

“We got the assets’ current and future potential valuation, almost as soon as we typed the address in our system,” said Skyline co-founder and CEO Guy Zipori. “To give us the valuations, our machine learning algorithm checked more than 130 data sources and analyzed the information covering the last 50 years.”

The real estate investment technology’s system concluded that the assets were listed at 12 percent lower than their estimated market value and reported the rental income at a significantly lower rate than comparable buildings in the area. It also showed similar buildings that got higher rentals after renovation.

Next, the AI technology suggested upgrades that could add value to the buildings the company was targeting, and calculated returns Skyline AI could expect if it made these investments. It’s no surprise that Zipori and his partners were thrilled.

“Within 6 hours we had booked our flights to visit the facility,” Zipori said, “and we committed to the opportunity in less than 24 hours.”

For him, the reason for excitement was obvious. Compared to traditional methods of analyzing real estate deals, Zipori had “bagged a great deal because our algorithm saved an immense amount of time and provided accurate analysis.”

Using AI to Drive Investments

Since the initial deal, the AI-driven platform has analyzed several real estate assets. According to Zipori, several deals are in the pipeline. While he declined to name the new purchases, he said, they were the first of many real estate deals powered by AI technology. Yet Skyline AI is not the only company using AI-powered predictive analytics to outperform current commercial real estate investment benchmarks.

Several tech startups, such as, AlphaFlow, and Ten-X are using machine learning algorithms to sift through listing services, match investors to properties, and find mortgages to purchase. Venture capital firms seem happy to pour funds into emerging AI-powered real estate startups, too.

Skyline, for instance, raised $18 million, while raised $5 million in Series A rounds. “AI is the future of real estate analysis and underwriting and is poised to fundamentally change the way assets are acquired,” said Haim Sadger, partner for Sequoia Captital, Skyline’s lead investor.

“These might be of different sizes or might be located far away from each other, but the AI software can find hidden correlations and generate a great deal of insight concerning the future performance of one asset based on the performance of the other.”

– Guy Zipori, co-founder and CEO of Skyline AI

According to Zipori, what makes AI analysis so bullish in real estate is the emphasis on predictive analytics. “One of our biggest advantages is that we have the most comprehensive data set in the industry,” he said. Skyline’s AI platforms mine data from over 130 different sources, analyzing over 10,000 different attributes on each asset for the last 50 years. Powered by natural language processing and high performance data infrastructure, all data is compiled into one large data lake, and then cross-validated to make sure the data used is accurate.

Machine learning (ML) algorithms, he said, are another big advantage. Since ML algorithms process data without any predetermined rules and continue to learn as they process more data, they analyze otherwise overlooked attributes and conduct multi-dimensional comparisons between seemingly unrelated assets. Conventional analysis models, he said, are much slower and ignore a lot of important information because they don’t have the right tools to consider the vastness of relevant data.

Typically, they would set up a peer group of assets that are of similar size, age and location. The idea is that similar properties in the same location will sell for roughly the same price, and move up or down at the same time. Skyline’s algorithm digs much deeper and creates a new concept of neighborhoods — clusters of properties deemed similar according to thousands of different signals in the data that traditional analysis might miss.

“These might be of different sizes or might be located far away from each other, but the AI software can find hidden correlations and generate a great deal of insight concerning the future performance of one asset based on the performance of the other,” Zipori explained. These new models, he emphasized, can be very useful when you want to invest in a new location or if there are no competing properties in the area.

A New Catchphrase

Zipori didn’t have to struggle much to sell the idea of predictive analytics to commercial real estate investors.

“What works best is to input the address a potential investor has current or recently sold and show them information they didn’t know themselves,” he mused. “Or [we] type the address of a property they recently bought and instantly generate a business model to improve returns.”

Encouraged by the response, Skyline AI now plans to expand beyond multifamily properties and invest in other assets such as office, industrial, and retail. In the process, the startup might change the real estate’s age-old mantra of ‘location, location, location’ to ‘data, data, data.’