Broadly, the commercial real estate industry has an inferiority complex when it comes to new technology. In some cases, this is unwarranted: US multifamily operations, for instance, outpace most other industries when it comes to AI adoption.
In other cases, however, the sense of inferiority is more than justified. Take the CRM, for instance. In other industries, AI-powered tools from CRM platforms like HubSpot and Salesforce have taken the data entry headache off sales teams, automatically logging interaction data and conversations on representatives’ behalf without hours of manual data entry.
But for a number of reasons, real estate – particularly the transactional side of CRE – has been slow to embrace these tools. This has been a major headwind for real estate services firms and deal sponsors alike, limiting the effectiveness of CRMs and firms’ abilities to analyze and get value from their own data.
Today’s letter will tackle the world of real estate transaction data through the story of one company attempting to crack the code: Lev. Initially founded as a marketplace for real estate debt, Lev has instead found success solving the sales data problem with a new kind of real estate CRM.
Specifically, we’ll explore:
- The real estate data landscape – and how it differs from that of other industries;
- Lev’s story and evolution;
- The evolving data landscape and its impact on operators and brokers alike.
Let’s dig in.
A Data Landscape
Anyone who has managed a sales team – in real estate or otherwise – is painfully familiar with the challenge of getting salespeople to enter information into a CRM. Plugging contact information, call summaries, deal status updates, and the like into an outdated web form is dull for anyone, but it’s particularly abrasive to the kind of people who are good at sales. Personality-wise, those who find joy in taking clients out to steak dinners are unlikely to be good at data entry.
Outdated and incomplete CRM data presents a number of problems that span industries including real estate. One, forecasting is only as good as the data it is based upon; if deals aren’t accurately represented in a CRM, it’s very difficult for a business to forecast future revenue or production needs.
Two, poor CRM practices mean that key relationships and context walk out the door when salespeople leave. If conversations and deals aren’t documented, it’s extremely difficult for a new salesperson to pick up where a past one left off. Deals die, prospects are left in the lurch, and revenue drops.
Some sales managers attempt to solve this with a carrot-and-stick approach, rewarding salespeople who do a good job keeping the CRM updated and punishing those who don’t. But that can only go so far; commission-driven salespeople tend to be motivated by dollars far more than managerial approval, and financially punishing salespeople for the crime of not updating a CRM is a bridge most managers are unwilling to cross. And it’s not lost on salespeople that keeping key information out of the CRM increases their negotiating leverage with their employer.
But over the past few years, AI has begun to fundamentally change this dynamic. Rather than relying on manual data entry, CRMs can now use AI to turn interaction data – calls, emails, calendar events, et cetera – into structured deal and prospect updates. Salesforce’s Einstein tool, for instance, takes advantage of the platform’s thousands of integrations to gather CRM-relevant information from a wide variety of sources. Other generalist CRM platforms like Hubspot, Attio, and Pipedrive have similar functionality.
Unfortunately, commercial real estate – the transaction side, specifically – has been slow to embrace the full power of AI in CRM, creating a struggle for deal sponsors and real estate services firms alike as they manage relationships with lenders, investors, and brokers. There are a number of reasons why real estate has been slow on the draw, but the biggest is that real estate dealmakers are generally more independent than the typical SaaS salesperson, with their own processes, preferences, and even CRM structure.
This makes embracing new technology at a firm level challenging, and it’s compounded by CRE having its own unique workflows and data structures that aren’t easily modeled in existing CRMs – creating an opportunity for a solution tailored to real estate.
The Lev Story
Lev may already be a familiar name to some Thesis Driven readers. The company made waves earlier this decade as a marketplace for CRE debt, raising $170 million in 2022 to build out its platform. But Lev had a unique approach to the marketplace model that set the stage for its next iteration: facilitate deals by working with debt brokers rather than replacing them.
“We realized that embedding within the existing ecosystem of brokers is the right way to build,” explains Yaakov Zar, Lev’s founder and CEO. Envisioning itself as a tech-enabled brokerage, Lev sought to incrementally layer technology and automation onto its existing network of debt brokers. “We wanted to start with technology doing 20% of the work, then go to 50%, and eventually make it to 90%,” says Zar. “If we could build a brokerage that was technology first, we could continue to iterate on the software until we were fully tech-enabled.”
The company was on track to originate $2 billion in loans in 2022 when rapid interest rate hikes put a chill on the market, dramatically slowing Lev’s business overnight. “We took the opportunity to re-evaluate what we were doing,” says Zar. “We had great internal tools, but borrowers don’t care about our CRM – it doesn’t win deals. We were a great technology company but a mediocre brokerage.”
So Zar decided to remake Lev entirely, running a tried and true pivot playbook: take your great internal software and turn it into your core product.
Lev’s internal CRM worked; it was capturing the information that brokerage firms otherwise weren’t seeing. Rather than build a brokerage from scratch, Zar reasoned, why not sell that software to those who needed it – brokerages as well as deal sponsors and others managing equity and debt relationships.
This kind of tech pivot has precedent: entrepreneur Stewart Butterfield was running a game development company before deciding to bring the company’s own internal communication tool to market. The game failed; the tool became Slack, which reached 12M+ DAUs before selling to Salesforce for $27.7B. And Tobias Lütke was selling snowboards before bringing his company’s custom e-commerce platform to market as Shopify, which now does over $7 billion in annual revenue. “Sell the awesome thing you built for your own team” is a proven playbook.
Zar built Lev’s CRM upon the lessons of Salesforce and Hubspot’s AI agents but designed with a real estate user in mind. “We looked at the detailed workflow of our users,” explains Zar. “Reaching out to lenders, managing lender data, getting deal feedback… all of those are unique workflows.” Like Salesforce’s Einstein, Lev uses AI agents to gather information across a user’s touchpoints – calls, emails, documents, and more – structuring and aggregating it.
“If a lender submits a term sheet, your CRM needs to parse it and get it into the system,” he continues. “Without AI, brokers would only add the winning bid to the CRM, which meant that firms would miss out on key data points that could help them understand the market. If you go from one to seven term sheets per deal into the system, that data moat you pitch as a brokerage is much more real. Your insights can come from real transactions, not just your research team.”
While LLMs are naturally good at the kind of document abstraction Zar describes, getting aggregate deal data into the kind of structure legible to real estate firms requires a custom approach. And Lev’s platform also brings in relevant information through, for instance, a CompStak integration and a database of more than 7,000 lenders. “So we can recommend specific lenders a sponsor might want to talk to on a given deal at the right time,” notes Zar.
Having multiple teams on one tracking platform makes coordination easier. “Finance teams, capital markets brokers, and investment sales teams are all tracking the same types of files and working on the same deals,” says Zar. “We’re offering them different panes of glass into the same transaction.” It’s worth noting that Lev’s seeing success with this new model, counting real estate firms like Time Equities and AANDAR Real Estate Capital as clients.
But how will increasing data transparency impact the real estate market?
Toward Transaction Data Ubiquity
Given the pace at which AI and data technology is evolving – and firms are embracing it – it seems inevitable that transaction data will become far more ubiquitous and robust than it is today. That is, the real estate market will have a much better sense in two years of the “market” for real estate debt and equity than it does today. While Lev is at the forefront of this market today, it’s part of a much larger shift toward data transparency that stretches well beyond real estate.
It seems inevitable that all this data will make the real estate financial markets more efficient. When the “market” for a deal is more transparent, truly “off-market” transactions in which one side is leaving meat on the bone will become rarer. Debt, which is more of a commodity financial product than equity, will surely get there first – so it’s not surprising that lenders are the bulk of Lev’s initial target market.
As data grows in importance, firms that accumulate lots of proprietary data – large real estate services players like JLL and CBRE, for instance – become more valuable, at least to the extent they can capture and hold their own data through a tool like Lev. Third-party data providers such as CoStar and CREXI without direct connection into brokerage CRM systems will likely suffer, as will sub-scale brokerages unable to compete with the behemoths’ data advantages.
But perhaps the biggest question is the role of the broker itself. In a world of transparent data on lenders, sponsors, and deals, do brokers still have a role facilitating transactions when far more cards already sit on the table?
Zar, for one, believes in the human broker’s enduring importance, seeing data as an enabling tool rather than an outright replacement.
“Anyone who has worked on a lot of deals has met brokers with savant brains. Two phones, everything in their heads. But most people aren’t like that; they need tools to do their jobs. If you can surface the right information at the right time, you can see the data and the right people to talk to all in one place.”
But real estate is still very much an industry built on trust and human relationships, even if those individuals are effectively meat wrappers on data models and LLMs.
“You still need someone to go meet the client and tell them why they should sell their property.”
-Brad Hargreaves