By Joshua Burd
Faropoint has found its place in the industrial real estate sector since its founding in 2012, with its focus on small to medium-sized warehouses in urban infill locations across the East Coast, Midwest and other high-barrier construction markets with strong demographics.
There is a catch, according to the firm’s Itay Ron.
“One of the things that we saw is that, of this trillion-dollar industry, 76 percent of the universe is either owned by owner-users or private owners,” he said. “So these are extremely fragmented with siloed pockets of information. They do not share information with one another. This is not sitting in a couple of brokerage houses that, if only they spoke, we would have all the information in the world.”
Faropoint’s solution was none other than artificial intelligence. Itay said as much during a presentation at NAIOP’s I.CON East conference in early June, where he detailed the firm’s push to aggregate and refine data from across its target markets and use A.I. to generate new insights on properties, improve its underwriting and support a platform that has acquired more than 20 million square feet.
The technology is a work in progress and far from perfect, he cautioned, while it’s meant to augment one’s business instincts and real-time conversations with professionals. But he’s confident that A.I., which has captivated Wall Street and society at large in 2023, is primed to disrupt commercial real estate in the years to come.
“It’s not a tactical move. This is a tectonic shift of the way we operate in our lives and the way we conduct our work,” said Ron, the Northeast market officer for Tel Aviv-based Faropoint, whose U.S. headquarters is in Hoboken. “It’s going to give, I think, a competitive advantage to whoever knows how to use it the best.”
The A.I. revolution is years in the making, he noted, long before the chatbot known as ChatGPT became a cultural phenomenon early this year, unleashing an arms race in Silicon Valley and a boom in technology stocks. Ron pointed to industries such as media, insurance and others that have been enhanced by the technology, from A.I.-generated news stories to the use of chatbots to process accident claims.
Commercial real estate is on the doorstep, he said, with its troves of data and a growing trend toward digitization.
“Commercial real estate has gotten to the point that both sides of the coin are ready,” Ron said. “We are all aggregating data way more than we were five and 10 years ago. We are getting to the scale that you need to say something smart about the world using these machines, and the machines are getting better in and of themselves. So by combining those two things, I think now is the time that we’re reaching an inflection point where these things are going take a much more central role in the way we operate.”
Digitization is step one, but a business must then collect and clean that information by “(putting) everything in a database in a way that’s coherent, in a way that has minimal errors,” before building a platform that is advanced and knowledgeable enough to be useful, he said. Equally critical is “the ongoing maintenance and management,” such as testing and benchmarking to ensure that the model is working and providing reliable, fundamentally sound information as part of a user interface that is accessible.
Ron, a senior vice president with Faropoint, also spoke to common pitfalls that are likely to emerge as A.I. spreads to real estate. That includes the concept of “garbage in, garbage out,” he said, citing the danger of collecting comps but not accounting for variables such as gross versus net rents, building heights, tenant improvement allowances and leasing commissions.
Attention to detail is even more important in today’s market.
“Especially in the past three, four years, we’ve seen how quickly information becomes stale,” Ron said. “From a year ago, there’s not a ton you can say about the world today.”
Faropoint, for its part, has leaned into technology since its founding. Ron noted that the firm collects market data going as far back as 30 years, which it combines with real-time data from its own network of brokers and business partners and then marries with so-called predictor data such as gross domestic product, household incomes and capital flows. That feeds a proprietary A.I. platform that “gets better over time, because it relies on more, larger amounts of data.”
“We train the model, and that basically creates a better user experience for everyone,” he said. “If we can provide instant feedback, we can be more competitive, we can give better pricing … The more data collected, the better the prediction. The better the prediction, the better the ability to underwrite and compete and then collect more data.”
There is still much work to be done, Faropoint executives say. Collecting and cleaning data is a perpetual task, as is the process of refining its A.I. platform. For now, Ron said the firm remains focused on making sound, fundamental decisions as “real estate people first,” while treating the technology as “more of a second opinion” that can help inform its strategy.
That could ultimately change.
“We are not there yet,” he said, “but if you get to a point where you trust it fully, you can provide almost instant feedback and instant underwriting on ultimately any building within a fraction of a second.”
