The opportunities presented by artificial intelligence (AI) today seem unlimited. During the early innings of this exciting new technology, the phrases generative AI and machine learning are often being used interchangeably. In the multifamily industry, practical applications of AI are primarily based on machine learning, a subset of AI that uses algorithms to learn automatically from big data to identify patterns and make intelligent predictions.
A multifamily leasing office becoming a screen with a virtual assistant that can mimic and go beyond human capabilities is not yet a reality in most communities. Nonetheless, our industry is identifying areas where true generative AI solutions can improve the resident experience and streamline operations. There is significant investment being made in developing AI solutions that are seamlessly incorporated into property management and accounting systems.
Multifamily owners and operators, drawn by the possibilities of finding efficiencies from adopting new technology, are testing various machine learning solutions that may eventually lead to true decision-making without human interference. At Bell Partners, we maintain a disciplined approach to applying machine learning solutions to our operating platform, and we find success in some areas while understanding that applications in other areas of our vertically integrated platform still need more time to develop.
What follows are three areas of our operations where we see real benefits from machine learning applications and opportunities for true AI solutions in the future.
1. Leasing
Our entry into the world of AI began in early 2020 with testing of machine learning applications in our leasing operations. After a pilot program, we deployed a virtual leasing assistant technology focused on the prospective resident. This assistant guides a prospect through the leasing journey in a seamless process, from answering initial questions such as lease rates and availability, through directing the individual in scheduling an on-site tour, to choosing a move-in date. We’re now using this virtual leasing assistant across our entire portfolio of 85,000 apartment homes.
During the pilot process in early 2020, we were naturally concerned about potential negative reaction from prospects, from our leasing teams, or from both in interactions with a virtual assistant. That concern proved to be short-lived. Because the people involved receive complete, accurate, and timely information from the virtual assistant; and because all interactions are in written form, with text that is naturalistic and easy to understand, we find that some of our prospective customers ask for the virtual character by name to speak directly with it.
For our leasing teams, our leasing assistant technology enhances customers’ experience by giving them more time to provide personal-touch service. Initial adoption was challenging but, over time, our teams appreciated working in tandem with the technology as a companion. For example, in cases when the virtual assistant is unable to fulfill a request, it seamlessly notifies a site associate in the local leasing office who can help. This protocol creates a flag in our system for our onsite teams to see the transfer and preserves the opportunity to maintain the prospect journey with us. The convenience provided to our onsite teams and prospective residents is invaluable, as it gives them an entity that can answer basic queries and provide quality assistance without regard to time of day, home time zone, or the leasing team’s availability.
The efficiencies being achieved from adoption of this virtual leasing program provide owners true peace of mind, and the program is generating real results overall. More than 72 percent of all Bell Partners’ prospect resident leads in 2024 have been supported by the virtual assistant’s 24/7 availability, with more than 80,000 of those leads being supported after business hours. Furthermore, we estimate that onsite teams were saved from fielding more than 96,000 hours of work—labor that the virtual assistant now supports, freeing up time for the staff to focus on turning ordinary interactions into extraordinary experiences.
2. Learning and development
Bell Partners is in the early stages of using AI in the learning and development space. Our learning and development team has used two machine learning programs with great success. Our Certified Bell Maintenance Technician program, started in 2022, enables our maintenance teams to accurately train for various real-world scenarios in a simulated 3-D environment via a computer or a VR headset.
The simulated sessions, including ones in which the user is repairing live electrical components, have been so positively received by our staff members that we believe it has reduced overall maintenance staff turnover by 14 percent. It has additionally increased the retention rate for those who participate in the program versus those who do not by an impressive 60 percent.
We also actively use a role-playing tool that was integrated into our training programs at the beginning of 2023. This program allows users to interact with virtual avatars controlled by live improv actors, thereby helping to troubleshoot such common workplace interactions between employees and their managers as performance feedback sessions, diversity and inclusion conversations, and conflict resolution. All our managers—a total of 600 associates—have been trained on the program with overall positive outcomes.
3. Research and data analytics
Our market research team has been another beneficiary of proprietary big data and advanced algorithms. Using anonymized address data gathered during prospect resident screening, the team was able to discern real-time migration trends in Bell markets and thus to gain the unique insight that many resident moves are indeed local (between cities or within the same city).
Gleaning such insights has practical implications such as reallocating marketing dollars toward migration hot spots and letting us focus future acquisitions within those nodes. As this example illustrates, AI and big data are important in the research field, as teams are well-equipped to mine large datasets and identify inefficiencies at the local level that can contribute to outperformance for both existing and new assets.
What may be next
Presently, those multifamily owner/operators who have adopted AI/machine learning tools for their platforms are now focusing on consolidating their applications to achieve a comprehensive solution, thus freeing up resources and facilitating change management. Although we have been cautious on our AI journey, we see a bright future ahead, as young renters who constitute the majority of our customer base are more likely to adopt a personalized experience with AI given their increased familiarity with technology.
Advancements in ChatGPT-like experiences for applications in leasing, resident management, maintenance, payables, data analytics, and the like are being made by many of our vendor partners. Predictive maintenance models that suggest when to replace equipment, intelligent lighting systems that optimize the use of energy, and resident screening applications that reduce fraud are all areas where AI solutions are emerging or remain theoretical (and years away from deployment) but nevertheless are attracting investment capital by vendor partners in our industry.
Despite this optimistic outlook, some hurdles lay ahead. Traditionally, the real estate industry has often lagged behind in adopting new technologies. Much of the data to develop these solutions already exists in property management systems, so more time and investment will be needed to make them a part of the digital ecosystem. Data privacy for our residents will remain an ongoing concern in any transactional environment. At present, many AI programs are not feasible, as implementing them would require putting a large amount of confidential data at risk.
For the multifamily sector, our experience suggests that great benefits can come from adopting AI and machine learning solutions and applying them to the sector’s operating platforms. For software providers, the AI arms race is just beginning, and there’s still much to do to get viable solutions for our resident customers. With a disciplined approach, we believe the AI evolution offers real possibilities to help us deliver superior customer experiences for our residents in the years ahead.