4 Takeaways From CREtech 2023

At every tech event this year, AI is being discussed like it's the internet in 1997. Last week's CREtech, billed as the built world’s largest innovation and sustainability event, was no exception.

But AI, or artificial intelligence, raises fundamental questions for even the savviest real estate decision-makers, especially during a time (unlike the ‘90s) when they’re faced with immense capital constraints.

Thought leaders at the Sept. 19-20 event in New York City may not have had all the answers. But the major themes that resounded in presentations and panels, from which all quotes included below were taken, certainly supplied the real estate community with plenty to ponder as the industry grapples with a new era of technological evolution.

When workers do return to the office, they expect their experience to be seamless — especially those in the digitally native millennial and Gen Z groups who will collectively make up 60% of the workforce in just a few years, according to Mary Clark, the chief marketing officer for cloud-based smart building firm Brivo.

That’s a big way technology shows up. “We're looking to make entry into the building as frictionless as possible,” she said.

At 93-story supertall skyscraper One Vanderbilt in Midtown Manhattan, SL Green Realty’s President Andrew Mathias explained, “The whole building is run by an app.” For example, he continued, employees don’t need a security badge — all access control is done through the app, through which they can also order food from restaurants on the site as well as get specials and promotions from neighboring businesses.

Notice how he said “one app.”

Several event attendees told LoopNet that there are simply too many competing and overlapping “solutions” in proptech, a market desperately needing more integration. “I think we're going to see an evolution more towards this single sign-on approach,” said Corinne Murray, founder of workplace consultancy Agate. “Whether they're coming from different vendors or not, there's got to be a single skin that every user interacts with.”

It’s partly those types of apps, which were demoed and discussed in droves on the event’s expo floor, through which decision-makers collect data. And if there was any word uttered more frequently than “AI” during the event, it was “data.”

The two are becoming symbiotic, though, and Starwood Capital Group’s SVP Betsy Reed said it best: “Data is the key, right? You can't do AI without data.”

The good news, perhaps, is that the industry has plenty of it. Commercial property professionals are collecting data at the customer level through “tenant experience” interactions — smart building apps, access control hardware, net promoter score (NPS) surveys and even natural language processing of customer interactions — and at the portfolio level through massive arrays of market analytics data.

But then what?

As commercial real estate companies get more serious about technology, data and artificial intelligence, they’ll likely dedicate more executive roles to it, similar to that of JLL’s Yao Morin, who is billed as the brokerage industry’s first chief technology officer hired to focus on AI. Explaining how she landed the relatively new appointment with just a tech background and no CRE experience, Yao said that JLL leadership came to her saying, “‘We have so much data, but we don't have a great way of organizing it.’”

And that’s where AI steps up to the plate, various speakers agreed: to help make sense of the data.

“We're gathering all of this data and it's frankly too much to be used efficiently,” remarked Greystar’s senior director of smart building technology, Cris Kimbrough. “So, I'm bullish about AI.”

But where to start? ChatGPT has gotten the most buzz in the broader environment, and so that’s where many real estate professionals have also turned their focus. With chatbots’ natural language processing tools, decision-makers want to be able to ask simple questions about their markets and their end-users and have the answers spit out to them.

ChatGPT is just the tip of the iceberg though, according to Josh Panknin, the director of Columbia University’s real estate AI research and innovation department. And it’s only as good as the data put in. “We use ChatGPT, but it’s one of 200 machine learning tools that we implement,” he said. “It's like a hammer. It does very well on a specific task, but if you're trying to change a light bulb, a hammer is not the tool that you want to use.”

As AI tools are evolving, there’s no need to rush in and bet everything on one way it can be used such as chatbot implementation, for example. “AI is cool, but you try to boil it down to how you'll use it, and that's where I think a lot of us are struggling,” said Executive Vice President Doug Pearce of Waterton. He referenced a common sentiment heard at the event: there’s a lot of FOMO, or fear of missing out, going on with AI adoption right now.

But AI still has a long way to go to be honed into what the industry wants. “We’ve taken the operational stuff and we've got that down,” Pearce continued. “We've learned to walk. Predictive analytics is an area that we think is learning how to run. Can we predict delinquency rates? Can we predict how many of our residents will renew in the coming 12 months with any degree of accuracy?”

Right now, industry professionals need to go out on a limb and find out what’s best for them, what’s best for their sector, the market and the region they’re operating in, and even the individual assets. And they’ll have to be comfortable not knowing what the return on investment will be.

“I might automate a process and that's going to eliminate one job — so there's my ROI; I can make sense of that,” founder and CEO of Mag Partners, Maryanne Gilmartin, said. “But, for example, we're developing a system that allows us to predict markets six to 12 months in advance; that will give us a higher return relative to our competitors because we got into the market early, and it's impossible to quantify our benefit and our returns. We don’t even know how much it’s going to cost to build, because it’s not been done before, so you have no idea. But those are the kinds of things that would benefit a real estate company.”

There’s no silver bullet tech solution, Brivo’s Clark said. “We know that we have to integrate with so many different technology partners to deliver more value to the end users that we're serving. For the most part, that’s the industry’s approach. But I think we need to do a bit better.”

Siloes in the industry are impediments to anyone’s advancement, Clark warned, urging the industry to be as “open-API-forward as we can.” Open API, or public API, are application programming interfaces used in machine learning tools like ChatGPT that are freely available to everyone so that the tool can learn from myriad sources and provide a more universal solution.

And that seems to be the direction the industry is headed as it adopts more machine-learning tools. “Traditionally, we made money in real estate with idiosyncratic or asymmetric information,” said Jeffrey Bermann, general partner of investment firm Camber Creek. “If I knew something that you didn't, I'm going to be able to leverage that and make more money. Now we're talking about entering an era of transparency with the benefit of the entire industry,” which is the only way to solve some big industrywide problems, he said, such as decarbonization goals.

Murray later shared a similar sentiment. “[The industry] is on the same team in a way that I think is more pronounced than it has ever been.” And Gilmartin echoed that notion in a big way. “I'm here as a rallying cry to say that we cannot keep doing things the way that our grandfathers did in the built environment. We need to embrace technology and specifically AI. And also, if we don't start sharing our data, then I don't think you can create this open AI system that allows us to aggregate it, compile it, analyze it and push out solutions.”

But that brings up big security implications. “When you bring in tech, understanding the security aspects of data privacy is going to be critical for adoption,” said Don Watson, SVP of real estate and facilities management at Oracle. “You have to be prepared to talk through that and show how you are protecting identity and security based on what you know about it.”

When you put something into ChatGPT, you also put it in the public domain, warned Clint Osteen, senior director of IT for Granite Properties. “So that's the first discussion we're having internally around ChatGPT.”

“It's of paramount importance to us that we're protecting that data,” Greystar’s Kimbrough said.

With the built environment reportedly contributing 37% of global emissions, and more than half of that coming from HVAC, lighting and other property operations, according to Honeywell real estate product manager Dani Stern, it’s more important than ever to reduce buildings’ emissions footprints through operational tweaks.

And with increased regulations and incentives from federal and local government to hit carbon-reduction goals, more and more reporting will be required, Stern continued.

“Sustainability is going to be a big thing that's going to be impacted by technology and it’s going to be crucial to empowering people to understand how much energy they're using,” RXR's Chairman and CEO Scott Rechler said.

“If you have the data, you can start pulling levers and see what changes that it makes in the building operations from an ESG [environmental, social and governance] standpoint,” said Hines’ chief digital strategy officer, Ilene Goldfine. “So that's where we are now, is using that data to drive the decarbonization decisions we're making in the property.”

With the right data models gathered from building operations — mostly through a host of sensors on equipment — “AI can make thousands of micro-adjustments to the building and individual spaces based on predictions and models that can enhance or even replace the activities that any single building operator can do,” Stern explained.

“The problem is, without AI and advanced technologies, if you pull one lever, you reduce the other,” he added. “So, if I'm going to focus on occupant wellbeing and productivity, that’s great. But guess what? You will not meet your operational efficiency and sustainability goals [then], because you'll waste a lot more energy to achieve those. This is really what the models are trying to do: get you both at the same time, which means that advanced decisions and advanced modeling on the machine learning side have to be deployed.”

There’s also then a tension between building owners and tenants, Stern continued. “Who pays for what? Is the tenant paying the market rate? The landlord might reduce energy costs, but what if the tenant is still paying the energy bill? How does the landlord get differentiation and maybe even another revenue stream while the tenant gets the sustainability goals that they need, along with the savings? There's a pathway there, and it's great to see owners and tenants working together to achieve that.”