AI in Multifamily Real Estate
By: Joshua Eng
Introduction:
The AI “boom” in recent years has caused undeniable efficiency improvements in
sectors like healthcare, education, and even transportation. Multifamily real estate
(sometimes shortened as “multifamily”) has long struggled with many inefficiencies of its
own. Property management spends time screening tenants, turning over units, and
optimizing rent. Meanwhile, tenants have to wait during this leasing process, causing
stress before they even step foot into their supposed new home. With AI revolutionizing
nearly every industry, it might be time for multifamily to take its turn.
This article will explore how AI implementation has affected the multifamily real
estate industry, how it can affect the industry in the future, and how AI in multifamily can
affect general society.
Current Uses of AI in Multifamily:
Firstly, let’s take a look at how AI is already impacting multifamily. In terms of
property management, AI tools have recently come out that promise to make
management easier. Firstly, AI-powered chatbots can respond to prospective tenants
instantly. In addition, smart scheduling software has proven to automate rent collection,
maintenance requests, and showings, improving efficiency. On the financial side of
multifamily, AI tools such as Cash Flow Portal and IntellCRE have come out that can
underwrite properties. Models can now analyze historical rent data, occupancy trends,
market conditions, and, of course, cash flow to project apartment values.Case Study “Nova Park Apartments”:
To project AI’s future impact, we can analyze data from Nova Park Apartments, a
multifamily complex in Garland, Texas, containing 198 units. We will use data collected
from November 28, 2024, to July 9, 2025. From the data collected, we found that there
were 340 chat messages by 37 unique users, and four tours were scheduled. We also
found that tenants tend to appreciate the convenience and 24/7 availability of AI for
common questions and scheduling tours. However, many prefer a human follow-up for
more detailed or personal inquiries. To estimate broader implications, we will have to
assume that:
1. Time per interaction: Before AI, each message takes approximately 5-10 minutes
of staff time to respond and 24-72 hours for tenants to get a response. After AI,
both drop to approximately 0 seconds. For our extrapolation, we will assume
each message takes 7 minutes.
2. Property management: Staff get paid approximately $60,000 per year for
apartments similar to Nova Park.
National Extrapolation:
Since AI leasing chatbots can save about 7 minutes per message, with 340
messages, this saves property management staff about 35 hours over 8 months, which
is about $1,015 in labor costs. Now, if we extrapolate AI adoption to all 40 million units
nationally, labor savings alone could reach approximately $306 million annually. Labor
savings could increase past this number if we factor in paid onboarding time (2-4
weeks). With AI, apartments could also decrease onboarding expenses, leading toapartments saving approximately $3500 per property manager they would typically have
to train. Furthermore, if apartments similar to Nova Park adopt additional aspects of AI,
such as predictive maintenance, approximately $10 billion annually in operational costs
could be saved.
Limitations for broader implications:
1. 2. 3. Nova Park is a mid-sized property in Texas. Therefore, results may not directly
scale to apartments of different sizes or in different regions.
The data period is only 8 months. Therefore, seasonal fluctuations in
leasing/maintenance could affect results.
Extrapolation assumes prospective tenant behavior and AI speed to be the same
across all properties.
Unlikely Potential uses of AI and their effects:
Now that we have covered how current uses of AI have and could affect
multifamily, we can take a look at some more uses that are not available on the market
yet and how they could impact multifamily.
1. Predictive Amenity Budget Allocation
By implementing motion sensors within every amenity, apartments could track
tenant usage and then use AI to allocate budget towards amenities effectively. For
example, if Nova Park apartment’s tenants really liked going to the gym, then Nova Park
should allocate the budget towards new weights rather than a renovated pool. Thiscould increase tenant satisfaction and retention as well as save thousands on
renovation costs over time.
2. Psychographic Tenant Pairing
By taking data from tenants, such as sleep schedules, dietary habits, and even
sleep schedules, AI could effectively match tenants with roommates that they will be
more willing to room with long term. This would increase both tenants' satisfaction and
retention as well as save turnover costs per unit.
Future Implications:
Now that we have looked at both current-day and future uses of AI in multifamily,
let’s talk about how widespread AI adoption will affect the multifamily industry as a
whole as well as society.
Industry effects:
Widespread AI adoption could transform how property management and the skill
required to be employed in the industry. Necessary tasks like scheduling tours,
responding to tenants, and basic data entry could be entirely automated before 2030.
This could inevitably lead to a decline in entry-level positions and an increase in
demand for AI-skilled workers like AI System analysts. Overall, this will most likely lead
to multifamily owners opting for a reduced leasing staff.
Affordability and Investment Incentives:On a broader scale, the efficiency gain from AI can affect more than the cash
flow of real estate owners; it has the potential for rent increases to be more moderate.
With a decreased amount of payroll expenses, owners could reduce rents while still
maintaining the same margins. This could happen due to pressures from consumers
recently wanting more “human connection" over AI work, with AI-managed properties
potentially being cheaper in the near future. In addition to affordability incentives, AI
tools could allow investors to more accurately forecast cash flow with the use of
financial AI tools. Investors may also invest in more multifamily deals due to the
potential for higher cash flow from lower payroll costs. This increased investment and
affordability incentive could potentially relieve housing shortages in traditionally
high-demand markets like New York or Los Angeles.
Potential Downsides and Mitigations:
With AI integration, potential privacy concerns could occur with
“over-enthusiastic” adoption. For instance, going back to the earlier idea on motion
sensors tracking tenant movement for predictive amenity budget allocation, some
tenants may see the sensors as an invasion of privacy. In addition, even when potential
tenants are applying for a lease, AI is known to discriminate against certain racial
demographics. Lastly, many tenants may prefer the “human touch” to property
management. They want to be able to see another human face when they go into the
leasing office, not a chatbot. This desire for human touch could potentially lead to
“Anti-AI” apartments, where tenants who do not want to live in an AI-controlled
environment could live. However, these “Anti-AI” tenants would most likely have to paymore, since those apartments will inevitably lose out on the net income gain from AI
efficiency.
Global and Societal Impact:
Outside of the United States, AI in multifamily could impact developing countries,
government policy, and, similarly to the U.S., housing. In developing countries, where
property management is less experienced and formalized, AI could increase efficiency
to a greater extent than in the United States, where technology is a lot more integrated
within management already. In addition, although it is impossible to know the amount of
multifamily units in developing regions due to unstandardized data collection, housing
shortages are prominent within developing countries, such as those in Sub-Saharan
Africa and Southeast Asia. In addition, many property managers in these developing
regions still rely on traditional pen and paper forms of bookkeeping and often have
informal “handshake-style” agreements with tenants. By improving these countries'
access to technology and AI, it could digitize rent collection, tenant screening, and data
collection, which would legitimize the properties for foreign investment. With increased
foreign investment, housing could potentially increase, leading to relief for many
overcrowded cities in developing regions. Right now, housing deficits in Sub-Saharan
Africa are anywhere from 50-100 million units, and Southeast Asia is estimated to need
50 million units to meet housing demands. AI and subsequent foreign investment could
be a potential solution for these regions to provide adequate housing.
Of course, there are limitations to consider with this. Many developing countries
still have inconsistent access to technology, let alone AI. In addition, aforementionedprivacy concerns would be even more prominent in developing countries with looser
data privacy laws and less digital security. Still, the possibility of AI making property
management in developing countries on par with developed countries is exciting to think
about.
Conclusion:
AI in multifamily real estate has the potential to transform every aspect of the
industry, beyond the leasing chatbots of today. The data from Nova Park Apartments,
although limited, demonstrates how even basic technological increases can generate
significant savings in time and labor cost. When scaled to the national level, these small
increases compound into billions. AI does not just have the potential to help managers
and owners; however, it has the potential to lead the market into being more equitable
and accessible as property management becomes more accessible and new multifamily
development is incentivized by lower expenses. Alongside the industry, the workforce
will have to transition away from soon automatable tasks to become more technically
capable. Globally, AI could narrow the gap between developing and developed
countries’ property management. If, hypothetically, AI improves the efficiency of
multifamily by 10% per year, how long until the industry we know now is
unrecognizable? Is that a good thing? We will just have to wait and see.
Acknowledgements: This article was reviewed by Peter Cachion, research assistant for
Wharton Professor Geczy. Data for the Nova Park Apartments case study was provided
by Shonda Huber, Vice President of Sales at Apartments 247.