A View on Proptech Post-NAR Settlement
Leveraging AI to Disrupt the Brokerage Model
I’ve been a proptech founder (backed by Pear VC) for the past year, which has been an absolute roller coaster given the changes in the U.S. market. I’ve been compiling my notes on the residential market as well as problems and opportunities I have been noticing in the space. Personally, I think we’re at an inflection point for real estate, where we have the opportunity to completely disrupt the industry leveraging a combination of AI and an incentive-aligned business model to transform the real estate brokerage.
Note: Many of these views are highly contrarian and there are a good number of brokers, agents, and even proptech founders and VCs who have built their entire livelihoods around the current establishment that I’m criticizing below that will likely take offense to this blog. (Out of respect for a lot of great fellow founders, I’ve omitted names of companies that fell into certain pitfalls I’ve observed when studying the market, but email me at roshan.srin@gmail.com if you have any questions.)
Market Overview
Warren Buffett sold off recent holdings he had in D.R. Horton, Lennar, and NVR in Feb. 2024. What’s striking about these investments in residential homebuilding is that recently purchased shares in these businesses back in June 2023. This move is particularly striking since Buffett once said in a letter to shareholders: “If you aren't willing to own a stock for ten years, don't even think about owning it for ten minutes.”
As we have seen in the market, higher interest rates in the last 3 years have directly impacted 30-year fixed mortgage rates. While the average rate in has dropped below 7% in the last 2 months, interest rates are at an all-time high over the last 30 years since the Great Inflation of the 1980s.
This has made homebuying particularly unaffordable for Millennials and GenZ who often prefer to rent over buy in the current market.
This is all to say that any way to reduce the burden on homebuyers in dollars spent on transactions will have a significant impact on real estate transaction volumes going forward.
Problems with Real Estate Transactions/Brokerages
The March 2024 NAR Ruling has Confused Everyone
The official ruling and changes of a massive settlement against the National Association of Realtors (NAR) that came into effect earlier this year have confused buyers, sellers, and agents/brokers. Prior to this ruling, a seller would offer a 5-6% commission to their agent, and when an interested buyer chooses to move forward on the property with their buyers’ agent, the buyers’ agent would negotiate their commission on the deal directly with the sellers’ agent.
The new ruling now states that:
A seller must offer a separate buyers’ agent and sellers’ agent commission. If no commission is offered by a seller, then a buyer must negotiate the commission with their agent. Essentially, buyer brokers are now prohibited from advertising their services as "free" to clients.
Buyers will need to sign a buyer-broker agreement with an agent before touring a property which may also include a fixed commission percentage in the agreement, which may or may not predicate the commission offered by the seller.
The impact is clear from an agent’s perspective. If no commission is being offered on a specific listing, they will not be as incentivized to show a home. Another clear impact is that commissions will definitely be coming down on the buyers’ side which means many of the 1.5M agents in the NAR will likely leave the industry.
Real Estate Brokerages have a false North Star
Because every real estate agent at a major brokerage is a 1099 contractor in the United States that pays a nominal “desk fee” for technology and training services to a brokerage, 90% of real estate brokers in the U.S. focus on one north star KPI: recruiting top agents as sales talent that can complete deals because every single brokerage operates on commission splits with agents - usually 60/40 or 70/30 in favor of the agent.
As a result, there is no incentive structure for brokerages to NOT RECRUIT bad agents because this is a numbers game to them!
In reality, the north star KPI for revenue generation is transaction volumes (their revenue is in fact a percentage of this), but unfortunately with the brokerage model, that’s been lost in translation.
Becoming a Buyers’ Agent has a Low Barrier of Entry
Most states have a Department of Real Estate which provides guidelines similar to California around who may be a licensed agent:
must be 18 years or older
Be a resident of the state
Not have been convicted of a crime
Complete 3 sets of courses
Take the real estate salesperson exam
Note that anyone with or without a college degree or high school diploma can become a real estate agent. While this may seem meritocratic, most agents who recently become licensed work as either part-time agents or lack an understanding of how to run a contractor business as a salesperson given that they are mostly on 1099s. This creates an environment where majority of agents are not “good” salespeople, and barely close any transactions at all.
Real Estate Brokerage Services are Separate from Legal Services
What is fascinating is that in most states in the U.S., a real estate attorney can also become a licensed broker directly. Yet, many real estate brokers that are attorneys either choose to focus on the brokerage model for offering services due to the historically high commissions they received, or legal services - but not both! Due to the nature of a transaction attorney’s work in verifying a transaction for a buyer and seller, many attorney-brokers can only be the attorney or broker of record in a transaction, but not both due to a conflict of interest. The following is the definition of a broker vs. attorney in this case:
Real estate brokers - Act as a neutral agent to facilitate communication between buyers and sellers. They write the purchase and sale contract, coordinate the transaction from the offer to closing, and help with closing for both parties.
Real estate attorneys - Ensure that the property transfer between buyers and sellers is legal and smooth, including title searches, prepare documents, review legal issues, and more.
In many states like New York, a real estate attorney is required to review a transaction at closing before a broker gets their commission.
Commission Structures are Misaligned
Buyer and seller agent commission structures are misaligned as stated above and always have been - especially on the buyers’ side. If I am a buyer’s agent and I make a commission on the sale of a home, I make more money when you spend more on a house. So I'm not as incentivized to help you negotiate a better price. Even if there is now a separate buyer and seller commission for agents, this doesn't solve the problem.
The Woes of Proptech Focused on Residential and Commercial Services
The Lifetime Value (LTV) of Buyers/Sellers is Low
As shown in the above chart, it’s easy to get new buyer leads interested in purchasing a property. However, the cost of acquiring the customer is incredibly high given the amount of effort that needs to go into showing them houses before a deal might potentially be closed. For every 500 leads a real estate agent gets, maybe 2-3 will close per year on average (worse in conditions with high interest rates).
Additionally, because most homebuyers or sellers complete a transaction every 5-7 years, the CAC-to-LTV ratio for these consumers is horrible in the status quo unless the buyer happens to be an investor in real estate and is a repeat buyer.
Most VCs have avoided Residential Tech Companies for a Reason
As a general rule of thumb, when I ask a VC firm why they have not invested in the residential space as often, especially around brokerages, there is often good reason and that’s because VCs have been burned by these companies before.
Historically, tech-enabled brokerages, fractional investing platforms, and SaaS services sold into brokerages and targeting agents such as recruiting tools, lead generation, and CRMs to name a few, all share the common problems of:
Having high upfront costs for initial customer acquisition (or recruiting talent as a brokerage)
High churn rates due to month-to-month subscriptions and no talent incentives for real estate agents
There is a Clear Lack of Product and Engineering Talent in PropTech
Completely anecdotal from my experiences talking to proptech founders (and founders generally in Vertical SaaS spaces), but I have generally noticed that majority of them come from a business background - whether it be consulting, private equity, or as a former real estate agent/broker themselves. This isn’t necessarily a bad thing, but it’s shocking to me that for starting a technology startup, many of these founders either lack a CTO, hire a CTO who was formally just a developer, or completely outsource engineering to offshore teams.
A large reason for this problem is that many of the top engineering talent in the U.S. from Stanford, MIT, UC Berkeley, Carnegie Mellon, U Michigan, UIUC, etc. have generally not been interested in proptech as a space, and prefer to work on problems in robotics, machine learning, and computer vision or simply to go work at a FAANG company due to their ability to make $250K+ in total compensation. Proptech as it is today is pretty boring for most engineers.
Obviously this creates a huge vacuum for most companies where good talent for product development is hard to find, which makes creating robust technology platforms difficult. As a result, many founders often focus exclusively on go-to-market, forming relationships with players in real estate and networking, but often at the detriment of building a scalable platform. This problem is the reason why many proptech companies run into CAC-to-LTV ratio issues at the Series A and Series B rounds, and often raise extension or down rounds.
Opportunities to Upend the Brokerage Model
Aligning Incentives between Buyers, Sellers, and Representation
Some brokerages are moving towards a flat fee structure for transactions. In terms of aligning incentives, that makes sense for buyers’ agents. However, a sellers’ agent should be paid based on deriving the best price for a seller which requires some commission. A hybrid structure to align incentives makes more sense.
Leveraging AI to Replace Back-Office Sales Work Completely
Just like how Uber used mobile apps and the creation of the App Store to disrupt the taxi industry, there is an opportunity to use AI to replace the back-office sales and transaction work being done by real estate agents and transaction coordinators.
A huge challenge in this space is how acceptable the use of AI is in a transaction and if it is possible for an AI agent to become a licensed real estate agent. I don’t know the answer to this question, but as a thought experiment to see if it is possible, I ran ChatGPT through 20 sample questions asked on the California Real Estate License Exam and it scored 95%. See the video linked here for more.
It’s important to note that AI is currently reactive and pulling data by scraping other sites to form an analysis, but if the accuracy is objectively better than the median real estate agent, there is an argument to license an AI assistant to assist in a transaction.
This includes everything from nurturing leads, which is what we did at my startup, to managing buyer-broker agreements to generating digital CMAs. It also includes having AI assistants manage all the parties involved in a transaction regarding payouts including a buyer, seller, escrow company, mortgage broker/lender/bank, title company, buyer and seller agents, and their respective brokers. (There is still an opportunity to keep one human in the loop who is licensed to validate the entire transaction process.)
Increasing the Standard for Representing Others in a Transaction
At risk of sounding pretentious, the educational barrier of entry to be dealing with high value products (homes in this case) is far too low in the status quo. I would argue that in hiring for a real estate brokerage, ensuring that all real estate agents are well-educated with a minimum of an undergraduate degree and can function like consultants or associates at law firms as opposed to sales reps would be far more useful. Some smaller real estate businesses do this today, but running a real estate business like a law firm requires capital upfront - arguably doable since this is how many law firms start out too.
Ideas of VC-Backable Opportunities to Disrupt Real Estate Brokerages (need further validation)
Zero-Agent AI Brokerage
Opportunity: Leverage AI to handle the entire brokerage process, from lead generation to transaction management, reducing or eliminating the need for human agents. This could lower operational costs and streamline processes since the human in the loop will be a broker, who can offer more services in a transaction.
Challenges: Ensuring the AI can handle complex, human-centric tasks like negotiation and emotional support. There’s also the challenge of gaining trust from clients who might prefer human interaction. (A potential resolution is to offer an in-person real estate consultant, who does not have to be licensed in the brokerage.)
Hourly Billable Representation (Law Firm Structure)
Opportunity: Offer real estate services on an hourly basis rather than a commission or flat fee. This could appeal to clients who need specific advice or services without committing to full-service representation. You can build an automated platform for managing time and billable accounting.
Challenges: Pricing transparency and value perception could be issues. Clients might be hesitant to adopt a new model without understanding the potential cost benefits. There also needs to be high enough volume for a set of common products with predictable pricing.
Brokerage to combine Real Estate Services with AI
Opportunity: Integrate AI to offer real estate brokerage, legal services, mortgage brokerage, and title insurance services all in one. This is essentially an AI-ified version of HomeServices of America.
Challenges: Integrating AI effectively requires substantial investment upfront and this would also be hard to scale with a need for brokers across multiple fields in each state to be compliant.
Investor-Focused Brokerage Solutions
Opportunity: Develop solutions tailored for real estate investors, such as tools for identifying profitable properties, analyzing investment potential, and managing portfolios. This could tap into the growing market of real estate investors.
Challenges: Operationally challenging due to the need to have multiple brokers in high ROI states for investors to work with.
Recruiting Platform for Quality Talent for Representation
Opportunity: Create a platform that connects brokerages with high-quality real estate agents and brokers. This could help brokerages find top talent more efficiently and improve overall service quality.
Challenges: Depending on what characteristics you index on, such as education level and sophistication, the pool of agents today is likely small and you would need to have a training manual to essentially convince high caliber management consultants to become real estate consultants.
Ideal Founders for these Businesses
I will admit that I am slightly biased here, but I do believe that having a product development background is far more important for building a technology startup than direct industry exposure, which can often be learned as you as a founder talk to customers in the space. That said, each of the businesses above require three core elements:
A founder with an understanding of operations - running a brokerage is first and foremost an ops business as opposed to a marketing or sales business. It’s easy to hire a marketing manager to help with SEO or recruit 1-2 agents to help with sales at a certain scale. You don’t need to be a broker yourself, but definitely an operator-type.
A founder with software or product experience - as a technology business, being able to build a scalable product is tantamount to the longevity of the startup.
Either a founder or an advisor with a broker’s license - this individual can be a voice of the industry to help with recruiting licensed agents, gaining deal flow, and relationship building.
Note that #1, #2, and #3 do not need to be 3 distinct people. It’s possible for 1-2 people to fill these roles (although finding a single person is incredibly rare).