The Next CRE Frontier: Recruiting Artificial Intelligence and Big Data to Boost Projects’ Performance|
Reading Time: 6 minutes
Alexa, get me our tenant mix recommendations.
OK, maybe it’s a stretch to think Amazon’s digital assistant will be providing commercial real estate insights anytime soon, but these days, more commercial real estate investors are turning to Big Data and artificial intelligence to help produce forecasts and valuable insights on property performance, long-term tenant behaviors, and future project development.
It’s about time the instinct-driven world of commercial real estate delved into so-called “predictive analytics” to help make better decisions, according to some experts.
Currently, most companies use only a fraction of the data available to them, and commercial real estate is “late to the game,” said Jeri Frank, with Cedar Rapids, Iowa start-up Stratafolio, an aggregator of real estate data.
“Engaging with your data can have a profound impact on company performance and provide an edge over others,” Frank told National Real Estate Investor.
Predictive analytics and other machine learning techniques will allow CRE firms to quickly and efficiently make connections between a wide range of data such as tenants’ financial reports, rent performance, retail customers’ mobile phone traffic, and local demographics to, for instance, optimize tenant mixes for a property.
By leveraging predictive analytics, developers can also reduce risks and make more accurate assessments of expected project values, site selections and market opportunities.
How Predictive Analytics Differs from Other Data Uses
To the commercial real estate professional, predictive analytics is still an all too new concept for now. It can be difficult to understand how this ‘new concept’ differs from other common data practices and uses of information technology. What strategic benefits can predictive analytics offer? Here’s a run-down of the various data strategies:
Descriptive analytics: This is the most commonly used analytics approach in commercial real estate, Joseph Ori, executive managing director at Paramount Capital Management, recently told GlobeSt.com. CRE firms use past data such as vacancy rates, rents and loan-to-value ratios to measure property and investment performance.
Diagnostic analytics: More advanced firms use the results from their descriptive analysis to look backward in time to help determine why a particular asset performed well or poorly. This technique is also in common use by the CRE industry, according to Ori.
Predictive analytics: Descriptive and diagnostic analytics are based on static data. But growing computer power and the rise of artificial intelligence now allows so-called “predictive” analysis, in which commercial real estate professionals can predict future events by combining past data as well as innovative uses of outside data – often in real time.
Prescriptive analytics: Even farther out, said Ori, is so-called “prescriptive” analysis, in which commercial real estate firms will be able to take the results from their predictive analysis to get practical advice based on artificial intelligence, such as picking which of several projects would be the best investment. Prescriptive analytics is in its infancy and won’t be common for at least 15 years, according to Ori.
The Next Big Leap
So for now, let’s concentrate on predictive analytics, the next big leap for commercial real estate firms. Ori predicts it will be in common use by the CRE industry within the next seven to 10 years.
Predictive analysis uses data mining, machine learning techniques and statistical algorithms to look forward and predict future outcomes, based on combinations of historical data and new types of data such as mobile phone location data.
Essentially computers learn from past behaviors and gobs of outside data — often in real time — to help you gain new insights and make better assessments on the outcomes and future success of your development projects.
A leasing broker, for instance, could use a publicly traded tenant’s financial reports — both numbers and text — to predict the likelihood the firm will renew or expand its leases, or run into financial trouble.
A report by market research firm Stratistics MRC states that the global predictive analytics market is expected to grow from $3.89 billion in 2016 to a hefty $14.95 billion by 2023, with a compound annual growth rate of 21.2 percent.
Tapping New Sources of Data
Some companies are already using innovative data to make such decisions.
Atlanta start-up Mogean has built a data science engine that processes location data of more than 150 million mobile phone users to provide its clients with “real life” behavior information on their customers. The data is generated when customers opt to share their locations with the mobile phone apps of retailers and consumer brands. Mogean is backed by Georgia Tech’s incubator program, the Advanced Technology Development Center.
According to Matt Reilly, Mogean’s chief executive, a “sideline” business at the company shows how such data could help shopping center developers and other commercial real estate firms.
For “one global coffee chain,” said Reilly, “we looked at where people who interact with the category (i.e. – visit coffee shops) move and commute, and helped them relocate stores based on the analytics. The analysis shows that it’s more important to watch where people are 30 minutes before and after they buy coffee than it is to consider where they live or work or (their) commute patterns.”
Such insights into customer activity could be a valuable service that mall operators and other retail property owners can offer tenants struggling to adapt to the changing retail world.
“It’s a dogfight right now for customer attention,” said Reilly. National mall operators could even offer tenants daily intelligence on how well promotions worked or how much customer traffic their competitors got, he said.
“It’s more than just a footprint thing,” said Reilly. “The question is, how can you help your tenants?”
Other industries and non-profit sectors, such as healthcare, education and law enforcement, are also leveraging such technology to improve their operations.
Atlanta’s Georgia Aquarium uses predictive analytics to help manage crowd traffic flows, monitor guest comments and predict future ticket sales.
Georgia State University, which has gained national attention for efforts to boost student graduation rates, uses predictive analytics to track risk factors for thousands of students each day. The program has improved four-year graduation rates by six percentage points since the program started in 2012.
More capital is also aiming to develop predictive analytics applications for the commercial real estate world, particularly among large-scale investors and leading commercial real estate firms. In the first half of 2016, investors provided more than $1.8 billion to real estate tech startups, according to CB Insights.
“Developers planning new projects have used our platform to virtually test different unit mixes and amenities packages to identify the optimal scope for their projects,” said Marc Rutzen, CEO and co-founder of startup Enodo. The Chicago start-up has launched a machine learning platform to produce key measures for multifamily properties, such as projected investment returns for upgraded amenities and occupancy and operating expense benchmarks.
Tara Bleakley, at Digital Map Products, suggests developers use traffic data to better understand activity in and around an area; geo-fencing can provide insights into demographics; and data on Uber and Lyft activity, paired with restaurants and parking, may help with decision making on retail site selection.
Predictive analytics can also improve the efficiency of commercial real estate firms’ operations by automating some chores. Instead of calling properties to survey rents, Enodo said developers can use its platform to automatically identify statistically relevant rent comps for their properties and projects.
Dabblers vs. Doers
To take full advantage of the technology, however, commercial real estate companies will need to build teams of people with the skills to harvest, process and analyze potentially massive loads of often noisy, incomplete and conflicting data, according to experts.
Those people include data scientists experienced in statistics and quantitative techniques, and industry experts who can help translate business issues into appropriate computer models and algorithms.
They’ll also need “data storytellers” who can present output in ways that decision-makers can quickly grasp – often as “heat maps” or other highly visual forms.
Building such a capability may sound daunting, but it can be developed over time, say experts. That includes partnering with start-ups or hiring consultants with expertise in the area and educating existing talent on the technology.
“If you are resistant to new technology, you’re ultimately going to be left behind,” said Rutzen. “Many CRE predictive analytics products are just starting to emerge and, although they’re not perfect, it would be better for developers to proactively familiarize themselves with these emerging technologies than to wait until their competitive edge has been eroded.”
Most companies use only a fraction of the data available to them to make key decisions, but a growing number of firms are using predictive analytics to help better understand and forecast trends.
Unlike descriptive analysis and other techniques based on past data from in-house and outside sources, predictive analytics uses a combination of artificial intelligence, past data and innovative “Big Data” sources such as geo-tracking of mobile phone users to develop new insights.
Commercial real estate firms are just scratching the surface of the opportunities presented through the use of predictive analytics, but a growing number of big players, private equity investors and start-ups are targeting promising uses for the CRE world.