Apartment companies use artificial intelligence to find insights in huge data sets.
Apartment operators and asset managers are gathering more data from their buildings than they have time to use.
The income from rents and the expenses from renovations – even the video feeds from security cameras – all contain information that might help managers and apartment investors improve or protect their properties, if they access the information in time.
Artificial intelligence (AI) applications can help. AI is capable of much more than helping students write papers on books they haven’t read. Even chatbots powered by AI are just the beginning.
AI is arguably making the biggest difference to the rental housing industry behind the scenes, in back offices where apartment companies increasingly use AI to make sense of all the data they collect and help them make better decisions about their properties – in time to make a difference.
“You’d have to be living in a cave at this point to not have AI touching your daily life as an apartment manager,” says Julie Flesner, Investor Relations and Communications Director for Investors Management Group (IMG), based in Los Angeles. IMG is one of many companies using AI to gather information.
Better audits through AI
AI applications can almost instantaneously process huge amounts of information and find patterns and correlations in the data, according to industry experts. For example, Landmark Properties, based in Athens, Ga., is exploring ways to use AI applications to help with audits of the company’s lease files and ledgers.
“There’s an opportunity to automate the identification of potential discrepancies and inconsistencies,” says Vanessa Young, Vice President of Management Service for Landmark Properties. The company is investing in AI systems that could improve its reporting and minimize the risk of financial losses due to incomplete or inaccurate records.
AI makes sense of data in different systems
Excelsa Properties, based in Bethesda, Md., is beta-testing a new system that uses AI to draw meaningful data from its portfolio of properties, which uses several different property management software systems.
“I’m able to standardize the data,” says Curtis Holder, Director for Excelsa. “Then I’m able to run my key performance indicators and analytical tools.”
For example, Holder is able to identify areas where Excelsa’s properties have gone over budget and find opportunities to improve occupancy and enhance other operations items.
Excelsa owns eight multifamily properties with about 3,000 apartments. It typically holds its properties for five to seven years while it performs renovations and improves the income. Its third-party property managers use a variety of software systems. “The data is not standardized,” says Holder. So, the properties in Excelsa’s portfolio naturally can’t use the same customer relationship management (CRM) to store all of its data.
Before Excelsa began using AI to bring this data together, that fragmented system created delays. At the end of the month, the property managers at the individual properties prepared reports. Then Excelsa’s asset management team of five analysts used the data to update their combined data set for the whole portfolio. “I have to go to Excel and I have to input the data,” says Holder.
Finally, the asset managers could use their carefully compiled data to analyze the expenses and income from the property.
“I’ve always had an issue in our operations with the fact that whenever we see financials they’re always a month behind,” says Holder. “It’s like managing in a rear-view mirror.”
It’s not unusual that the asset managers spot something in the data they would like to improve. Maybe the property managers went over budget on marketing. Or maybe turnover costs at the property are too high. “Well, you’re a month and a half past executing on that,” says Holder.
The new AI system allows Excelsa to get this useful information from its properties much more quickly. Information flows swiftly from the CRMs at the individual properties into a central database. “We get data in real time,” says Holder. “If I’m over budget, I’ll know it on the 15th of the month; I can start making changes to save my budget.”
The system Excelsa is beta testing also allows it to aggregate data such as expenses across its portfolio.
“Every operator has their own set of specialty analytics,” says Holder. “Getting the data into one location where we can allow our tools to work — that’s critical.”
For example, Excelsa has created its own tool to compare the cost of its renovations with the likely rental income at each property. With this information, Holder’s asset managers can carefully calibrate their renovations to the opportunities in the local market.
“We’ve got to make sure our dollars are accretive,” says Holder. Holder’s team carefully tracks the numbers for each individual property as the renovations progress and the local market changes. If any numbers move in the wrong direction, they make a decision to balance the equation.
When the information is available, perhaps from disclosures from publicly owned companies, the AI system compares these costs to the costs at competing apartment properties for items like paint. “If you are spending 10 percent more, then you can see that there is a potential savings,” says Holder.
The new system has already helped Excelsa avoid fines. One of its apartment properties is part of a larger, master-planned community and was late in paying the homeowners association fee for its units, racking up a substantial late payment penalty.
“It popped up through the AI as an overpayment of several thousand dollars,” says Holder. “The AI can scour the financials and say: ‘This is out of the ordinary – what is it?’” Holder’s team of asset managers alerted the third-party property managers at the property in time to avoid more fees.
AI helps with investor reports
The investor relations team at IMG is also using AI tools to draw insights from its data.
The task is simplified because all the properties owned by IMG use the same property management software system — so all the data is stored in the same centralized CRM system.
The AI quickly aggregates this data to create specialized reports that IMG’s individual equity investors expect. Each investment firm has its own reporting requirements. IMG currently owns more than 22 apartments properties with a total of 5,200 units. IMG typically plans to hold its properties for three to five years. This portfolio has more than 1,200 individual, active, equity investors, including accredited investors who have contributed small amounts through crowdfunding platforms.
Each gets their own, customized, monthly and annual reports on the properties they are invested in. In addition, IMG provides quarterly webcasts on the financial results and the status of renovations at its apartment properties through its investor portal.
“The apps are aggregating all sorts of massive data sets and transforming them into the snapshot pictures that ownership groups love to see,” says Flesner.
The app uses the “latest and greatest AI tech to aggregate very complicated data,” she says. That includes quickly differentiating between similar data — like physical occupancy versus economic occupancy — to create a consistent data set to build the reports that investors want.
To maximize the efficiency of its systems, IMG limits its portfolio to the use of one property management system, says Flesner.
When IMG buys a new apartment property, either the third-party managers at the property switch software systems, or IMG switches to a different management company. IMG has recently started its own management division, Parktown Living.
AI detects fraud
AI systems are also helping IMG’s third-party managers catch fraudulent rental application documents before these applications turn into signed leases and missed payments.
IMG’s properties in the Atlanta area have identified a significant number of fraudulent documents submitted by potential renters, from faked paystubs and bank statements to entirely invented identities. Atlanta is the No. 1 market nationally for leasing fraud, according to a recent NMHC survey of U.S. apartment operators.
AI apps can identify the tell-tale signs that a document is fake, such as differences in the fonts on bank statements that indicate that they were altered.
“The system keeps training itself to detect these irregularities,” says Flesner.
AI’s other uses
AI can also appear to converse. An AI chatbot can be trained to find patterns in earlier, successful communications with residents and potential renters to help it respond to new questions posed by live human beings. These systems can appear to improvise or even tell jokes. However, a significant number of customers who use AI chatbots still often eventually ask to communicate with a person.
“Their ability to handle general inquiries has improved, but they still struggle with more nuanced questions,” says Elie Rieder, CEO of Castle Lanterra, based in Suffern, N.Y. “Site teams are increasingly integrating AI into their daily workflows, but it isn’t replacing leasing agents anytime soon.”
Property managers increasingly rely on applications like chatbots to reach residents and potential customers more quickly than human leasing staff can respond – particularly after hours.
For example, Landmark Properties has significantly reduced the number of residents who are late in paying their rent. The overall delinquency rate has recently dropped 60 basis points in 2024 – a major improvement – after the firm began to automate delinquency follow-up procedures using AI.
“This success stems from the AI’s ability to consistently engage with residents, even outside of normal business hours,” says Young.
Other uses of AI are still in the future for the apartment business. Businesses such as casinos already use AI to instantly analyze the information from cameras watching their table games to identify fraudsters. Privacy concerns limit how cameras and AI might be used in apartment communities, but existing security cameras could be wired to AI applications trained to identify safety problems and immediately contact property management staff or first responders.
Bendix Anderson is a freelancer for units.