AI is poised to transform real estate across many fronts, with the potential to unlock up to $550 billion in annual value globally across the industry. For industry leaders, the challenge is to shift their AI strategies from experimenting with isolated tools to reimagining entire business domains with humans and agents working in partnership. How will this future take shape within the housing sector? In a conversation with McKinsey Global Publishing’s Eric Quiñones, McKinsey Partner Alex Wolkomir explains how housing leaders can transform their operating models and enhance customer experience with AI. An edited version of their conversation follows.
Eric Quiñones: How is AI changing the housing market beyond traditional property operations?
Alex Wolkomir: You need to look across the whole property value chain. AI is not only affecting how housing companies operate after a resident or buyer enters their system; it is also changing how people discover housing, compare options, get questions answered, assess affordability, schedule tours, and place their trust. This matters because housing has always been both a physical product and a “feelings” product. When people buy or rent a place to live, they are making high-stakes decisions about lifestyle, location, services, community, and other personal factors.
The front end of that journey is starting to shift. Traditional search, portals, broker websites, community pages, and property websites still matter, but consumers are increasingly incorporating AI-powered answer engines, AI assistants, and recommendation-driven discovery. For housing leaders, that means thinking about how to make sure their inventory, brand, policies, availability, pricing, incentives, service promises, and local knowledge are available to consumers wherever they begin. Operators, brokers, sales counselors, leasing teams, and service teams will need to move work forward faster and with fewer dropped handoffs. And they must ensure that consumers trust that the information they see is accurate, fair, compliant, and backed by humans who can deliver.
Leading companies will redesign the consumer experience from search to service and avoid the trap of just bolting AI onto existing workflows. AI also has the potential to transform how homes are built by making design and scheduling more efficient and shortening the time it takes to complete a home.
Eric Quiñones: What consumer behaviors should housing leaders pay attention to?
Alex Wolkomir: Consumers experience housing as a journey: for instance, they discover, compare, qualify, tour, decide, transact, move in, live, renew, or move again. AI gives consumers more options for self-service, instant answers, and personalized guidance, but not every consumer wants the same interface and interaction. Some who want speed and control may be comfortable with an AI-first experience if it is accurate and transparent. Some who want confidence and ease may like AI-assisted guidance but want preapproved options, fewer surprises, and a person available when the decision gets more serious. Others may prioritize discretion, reassurance, and a single accountable person. For them, the best AI may be invisible: helping the agent, leasing associate, sales counselor, or community manager prepare, follow up, remember preferences, and reduce friction without making the experience feel automated.
This is where the design of AI matters. In my work, I tell real estate companies that their approach to supporting consumers should be: “AI when you want it, human when you do not, confidence by default.”
The risk in the industry right now is designing one generic AI journey for everyone. That will feel efficient for the company, but not necessarily trusted by the customer. The better approach is to let customers decide when they want self-service, guided support, or a person.
This is especially important in housing because affordability, financing, concessions, repairs, warranty issues, resident complaints, homeowners association matters, and move-in problems all require care. AI can help make the experience faster and more consistent, but the human moments still matter. The goal is to make the experience feel more responsive and reliable and less fragmented.
Eric Quiñones: Where can AI create the most value across the housing value chain?
Alex Wolkomir: The biggest value will not come from a long list of use cases. This is not like playing Pokémon: You don’t have to catch them all.
Using AI to draft an email, summarize a document, or generate marketing copy may save time, but real value will come from redesigning full business domains. In housing, these span the full value chain: consumer and commercial domains, such as research and discovery, sales, and financing; delivery and operations domains, including construction scheduling, inspections, and community management; and enterprise domains, such as procurement, finance, and legal.
The highest-value domains are the ones where AI can shift from a tool to a system of actions. Isolated use cases do not have enough impact to create real value. On the other hand, trying to transform your whole enterprise with AI is too broad a goal. Domains are the ideal middle ground: They have one owner, one measurable outcome, a repeatable workflow, clear handoffs, and enough value to matter.
Eric Quiñones: How should housing leaders get started?
Alex Wolkomir: Start with the work—focus on operating models, not AI models. Pick one domain and think about questions such as: Where does the workflow trigger come from? Which systems does it touch? Where does the work slow down? What would improve the experience for the buyer, resident, homeowner, board, owner, broker, or onsite team?
Leaders can follow a relatively simple path:
- In the first 30 days: Pick one domain and one KPI, for example, shortening the time from triage to dispatch in maintenance. Map the workflow. Separate repeatable steps from activities that require human judgment. Define what can run, what needs approval, and what should not be automated.
- Over the next 60 days: Connect the workflow to your systems of record. Build human approval checkpoints into high-risk decisions. Measure your progress weekly based on operating outcomes.
- After 90 days: Develop reusable agents for common workflow steps such as intake, preparation, execution, follow-up, approval, and logging. Remove approval steps only in workflows where agents have proven their reliability.
And remember that becoming AI-enabled does not happen overnight. You have to be intentional about redesigning work in a fundamental way, including redesigning the roles you are hiring for on your teams. Impact isn’t just about deploying new technology; you need to proactively create a better overall experience for your customers and your teams.
Eric Quiñones: What are practical examples of AI moving from pilots to outcomes?
Alex Wolkomir: One of the strongest examples is the lead-to-appointment domain. Let’s say someone calls after hours to inquire about a housing community. In the traditional model, the inquiry may sit until the next business day. By then, intent has cooled. In an AI-enabled workflow, the system can answer approved questions, assess the consumer’s intent, book an appointment, create a record in your customer relationship management system, and prepare a handoff pack for the sales team. A human still owns the relationship, but the customer does not wait in a dead zone.
A second example is affordability and financing. In home buying, the customer journey often fragments between sales, mortgage, incentives, and underwriting. AI can help consumers understand affordability scenarios earlier, route customers to the right human expert, and reduce friction through the steps from the first expression of interest to a conversation about financing. This must be designed carefully because lending, affordability, and qualification raise important compliance and fairness considerations. But AI can provide speed and clarity that improves the customer experience.
A third example is pricing and incentives. AI can help operators and builders understand demand signals, inventory, sales velocity within a community, days on market, and the effectiveness of incentives. But governance matters: While AI can help develop insights and recommendations, leaders still need clear guardrails, appropriate legal and fair-housing compliance review, and human accountability for pricing and incentive decisions.
A fourth example is warranty or maintenance. A warranty issue or maintenance ticket includes numerous steps from intake to completion, including triaging, collecting evidence, assigning vendors or tradespeople to do work, communicating with homeowners or residents, coordinating property access, monitoring service-level agreements, billing, and understanding the root cause of problems to help prevent them from happening again. AI can help classify issues, ask for missing information, route work to the right tradesperson, draft communications, flag exceptions, and update the system of record. The common thread across these examples is that AI creates value when it moves work forward with the right controls.
Eric Quiñones: What risks should housing leaders keep in mind?
Alex Wolkomir: The first risk is pilot purgatory. Many companies now have AI experiments. Some are useful. Some are impressive. But if they sit beside the work instead of inside the work, they do not move the business forward.
The second risk is measuring the wrong thing. The important metric is not whether people opened the tool or used more tokens—it is higher sales and conversion rates, shorter cycle times, faster reporting, higher customer satisfaction, better first-time resolution, and fewer escalations.
The third risk is automating a broken process. AI can make a bad workflow move faster, but that is not a transformation.
The fourth risk is eroding trust and compliance. Housing is a regulated, relationship-heavy, high-stakes environment. AI cannot be used without human oversight to handle work related to fair housing, privacy, lending, pricing, resident communications, and other sensitive areas.
Leaders should think about a tiered approach to AI. The first tier would be fully automating low-risk, high-volume work, such as status updates, scheduling, FAQs, and document retrieval. The second tier would enable AI to take action within guardrails in areas such as drafting documents, routing work, creating handoff packs, and collecting evidence. The third tier would utilize AI but require human approval or a human in the lead for work related to finance, exceptions, legal or compliance issues, sensitive data changes, contract revisions, and high-trust customer moments.
The fifth risk is losing control of customer relationships. As portals, AI search, answer engines, and third-party assistants shape discovery, housing companies may have less control over how their offerings and brands are represented. The answer is to ensure your information is accurate, trusted, and easy for these channels to use, while building workflows so you can respond quickly and consistently wherever demand arises.
The final risk is losing your competitive advantage by using common AI tools in the same ways that other companies do. The differentiator will be how effectively you design workflows, use your proprietary data, build customer trust, and learn from every interaction to improve performance and provide better customer experiences over time.
Over the next few years, the companies that win with AI in residential real estate will be the ones that can turn demand into efficient, trusted execution at scale, without making the experience feel less human.
Related Articles
