Alex Shtarkman has spent the past decade doing something most real estate venture investors don’t: getting his hands dirty inside the operating businesses he backs. As Managing Partner and Chief Investment Officer at Antithesis Ventures, Shtarkman splits his time between investing and advisory work, giving him a front-row seat to how real estate tech companies actually perform once the deal is done, not just how they pitch.
That vantage point has made him one of proptech’s more consistent skeptics of conventional venture capital. Shtarkman has long argued that venture-style growth expectations don’t map cleanly onto real estate, where liquidity tends to arrive through sub-$100 million M&A rather than the billion-dollar outcomes venture is built to chase.
Shtarkman will moderate a panel at Blueprint in Las Vegas on September 22, “Non-Hype AI: Practical Applications Across Real Estate Asset Classes,” primed to be a candid discussion of the real-world applications of AI inside their organizations. Ahead of his panel, Shtarkman spoke with us about why the venture model has struggled to fit proptech, where he still sees genuine opportunity, and what he’s watching for at the conference. Anyone interested in hearing Shtarkman’s hot takes on proptech investment should join us in Las Vegas.
NP: What have you learned from working directly in the real estate industry versus being a passive investor?
AS: I’ve always been a hands-on investor. Almost like a hybrid of investor and operating partner. My experience has really been forged in the trenches, working alongside founders and management teams through all kinds of situations, good and bad.
I think it helps to start at the macro level and then narrow down to what’s happening in proptech specifically. Venture as an asset class has changed dramatically. A couple of years ago, I started a blog called “Unpacking Proptech” to deconstruct what was actually happening in the built environment at a time when I felt venture itself was fundamentally changing. I actually published a new post on this just this morning. If you look at seed, Series A, Series B, and venture growth rounds today, 80-90% of those dollars are now concentrated in $25 million-plus financings. That leaves roughly 10% of dollars for the average company across every vertical.
The venture world has split into two camps. A handful of mega-funds — the Andreessens, Sequoias, General Catalysts, Benchmarks — have become what I call “AUM machines.” Andreessen alone raised nearly a fifth of all LP capital in the first quarter of last year; 75 cents of every LP dollar raised went to just five firms. These firms invest with a generalist, power-law mentality: get as many bets into the funnel as possible and let the winners emerge. Very few firms can actually run that playbook. On the other end of the spectrum is a growing world of specialists. These are pre-seed through Series B funds with a defined underwriting edge, like cyber or deep tech. The “fat middle” of generalist funds is getting squeezed out. Top-quartile venture today produces about 1.5x cash-on-cash returns — worse than top-quartile private equity, on a much higher risk profile. And there’s been essentially zero liquidity across the last eight venture vintages, with fund lives stretching to 15-20 years. LP capital is responding by either pulling back from venture entirely or concentrating with the funds running that diversification-at-scale playbook.
Now overlay that with proptech specifically. I went back to a founder survey we ran three or four years ago with the Center for Real Estate Technology and Innovation. Only 13% of proptech dollars came from traditional venture firms. More than 80% came from strategics such as real estate angels, corporate CVCs, and proptech-focused funds. Ninety percent of those dollars came in as convertible notes and SAFEs rather than priced rounds—bridges upon bridges that weakened many of these companies’ cap structures. Seventy percent of founders said they wanted real estate strategics on their cap table, but after getting them, 70% said those investors added little to no value, and 85% of founders who raised from proptech funds felt the LP-network value they were pitched never materialized. And yet, asked again, 75% of founders still said they’d want to raise from a fund or corporate real estate strategic.
Four structural problems mirror what happened in broader venture: an underwriting problem, a scalability problem, a liquidity problem, and a conflicts-of-interest problem. On underwriting, a lot of these strategics invested because a technology looked interesting and could plug into their own real estate portfolio to drive ROI. That’s a fundamentally different psychology than investing to build enterprise value toward an exit. That created a fast-money dynamic around notes and SAFEs, where valuation didn’t matter much and FOMO took over, which fed directly into conflicts of interest. In proptech specifically, your main customer, your main investor, and your main acquirer are often the same person. If that same strategic investor mispriced your last round and you’re now short on runway, they’re also your most likely acquirer.
On liquidity, I’ve argued for years that venture liquidity skews smaller and more private than people assume. When I reran our data from three or four years ago, two-thirds of proptech liquidity came from M&A deals valued under $100 million — strategic and private-equity exits. Recent activity suggests that range hasn’t moved much, and I suspect the underlying problem has only gotten worse with time.
On scalability, things simply take longer in real estate. Every company shows up to its Series A with a marquee institutional real estate logo in the deck, but going from one building to five to 15 to 50 takes real time, and these companies don’t scale on a venture timeline. By the time they find their footing, they’ve often locked in customers on contracted revenue that may never fully materialize, against a small base of actual billed revenue. As funds have gotten bigger and shifted toward power-law investing, that’s a fundamental mismatch with how real estate companies actually grow.
Many of these businesses are, at their core, tech-enabled services companies with transactional or one-time revenue, not something that supports a 10-15x revenue multiple. Building a successful software, AI, or data business in real estate almost always requires a services layer, or an M&A roll-up strategy to acquire assets and locations. That’s a lot of hand-to-hand, operational value-add work, which is arguably why private-equity-style strategies are often more effective than traditional venture for early-stage real estate technology companies.
NP: Where does that leave early-stage proptech companies trying to raise capital?
AS: I see two profiles. First, companies that have already raised: many are stuck on cap tables built on notes and SAFEs, carrying valuations that were a relic of the post-COVID market. They’re trying to reinvent themselves with AI now, but legacy baggage makes it harder to raise again and harder to exit.
In some cases, their existing investors and board members have no capital left to deploy and have been sitting in the investment for five or eight years, deciding whether to stick around. My advice to those companies: if there’s a will, there’s a way. Many have built real IP and real revenue. They just need to rewrite their story, typically through a recap or repricing transaction that cleans up the cap table and balance sheet and effectively recasts the company as a “clean” seed or Series A/B asset.
Second: two, going on three, of my current advisory clients are $20-30 million-scale companies that are profitable and growing, and have never raised a dollar of venture or private equity (one has raised some debt). They’ve scaled through revenue alone and have now hit an inflection point. They have a software or AI story, or a plan to build a balance sheet and acquire smaller, stuck competitors with niche products. That’s increasingly where these companies find themselves, especially as the cost of engineering and compute keeps rising and they need capital to keep building.
My broader advice: raise as little as you can, at the most prudent valuation you can, until you know exactly what kind of company you’re becoming. Twenty-four to 36 months into a raise, you might think you’re a pure software or AI business and discover you’re actually an enabled-services company, or the reverse. Until that’s clear, preserve optionality. The real challenge isn’t raising $10-15 million today with a strong team and early proof points. It’s what happens 24 months after that raise, when the market’s definition of “success” has moved and you’re no longer seen as being on a path to a $500 million or billion-dollar outcome. But the reality is that most of these companies will be $100-300 million private equity exits, which are genuinely great outcomes that can create generational wealth for the people involved. The disconnect is that venture doesn’t count that as success.
One more data point that ties this together: overlaying that $25 million-plus financing concentration against graduation rates, roughly 80% of companies fail to graduate from seed to Series A within 24 months, and of those that do raise a large Series A or B, 75% fail to graduate to later-stage growth capital within another 24 months.
Run that math, and you get something like 7,000 companies stranded in the venture pipeline every two years. Attention and capital chase the flavor of the day — AI, right now — and momentum scalers on the venture J-curve. Proptech companies that often move from $1-3 million in ARR to $3-5 million and don’t have a visible path to $15-30 million for four, five, or six years get stranded in that bucket. The pool of capital that used to be available to them through angels, corporate VCs, and proptech funds has shrunk, so they’re getting squeezed from both ends.
NP: Let’s shift to Blueprint. What are you most looking forward to with your panel, and what do you hope the audience takes away?
AS: I spent a lot of time thinking about a topic that fits the day’s theme — obviously AI — but from a slightly contrarian angle. My relationships with real estate owners, operators, and investors have built up organically over more than a decade, and I’m constantly hearing what’s actually working for these organizations and what isn’t, where they’re investing and where they’re not, and trying to act as a diligence filter that surfaces only the opportunities that matter to them.
Honestly, I think people are getting a little tired of the nonstop AI narrative. Every headline makes it sound like everything is up and to the right, and every founder is asking why a competitor just raised $30 million when they can’t. So I wanted to put together a panel organized by asset class — multifamily, hospitality, retail, commercial — with operators I know personally who are actually implementing AI on the front lines and have learned the hard lessons about what works and what doesn’t.
This is meant to be a non-BS, tactical conversation. The statistic hanging over the industry is that somewhere between 75-90%, depending on the study, of enterprise AI pilots aren’t producing meaningful results. I want the panel to get specific: what’s actually ready for AI, what isn’t, where the puck is moving, how these operators are investing and managing risk, including the vulnerabilities AI has introduced into their operations.
The panel includes Eric Lamb, Operating Partner of Blackhorn Ventures, one of the largest general contracting firms in the world; Christian Lee, who just took on a significant new role at a major hospitality incumbent running a new division; and Mark Stutzman, who’s building an interesting retail-meets-experiential concept. We’re also working on a possible bonus addition from the multifamily or commercial real estate world.
NP: What’s the most impressive real estate AI use case you’ve come across?
AS: I’d point to hospitality and personalization. The category has evolved from traditional hotels to Airbnb-style stays to a newer wave of “apartment hotels,” spanning full-service to no-service, tech-forward operators. Access control, digital concierge, digital amenities, and feeding information to guests have largely been figured out. Personalized matchmaking against experiences, curating trips, recommending restaurants. That’s not new, but it’s getting better.
What’s changed is that AI has become an equalizing factor for the small operator, whether that’s someone managing a handful of Airbnb units or a boutique hotel, to deliver something approaching Four Seasons-level service. And ironically, it’s less about the technology itself and more about how it informs the human touch. A good PMS should be able to tell staff that a returning guest travels with a dog, or has a food allergy, or a specific entertainment preference. So when they check into a different property in the same portfolio, there are snacks waiting for the dog. Those small details are what create loyalty.
We went from an era of human service, to technology-powered service, to now technology as an equalizer that informs and elevates human service, particularly for smaller operators who want to look and feel like a major brand. It doesn’t have to be perfect. Simple things like digital concierge, digital access, and a quick answer to “I’m arriving at midnight, where can I still get dinner?” go a long way.
Historically, an independent operator’s tech stack ended with a PMS and a revenue management tool for pricing, maybe a slightly more advanced booking engine. Personalization stopped there. Now these platforms can empower in-house cleaning and maintenance teams to perform functions they never used to handle, elevating the entire guest experience.
NP: Other than the panel, what are you most looking forward to at Blueprint this year?
AS: It’s always a great event, partly for the familiar faces, but mostly for the new ones. Proptech has gone through a real cycle over the past five to eight years: some companies and investors have exited the space, and new ones keep entering. Pockets of resiliency are showing up in unexpected places. Modular and prefab construction, for example, went through a boom, a wave of failures, and a dry spell where nobody wanted to fund the category. Now I’m seeing a rebirth of modular and prefab companies attacking the market in new ways. What I always look forward to is walking away with one, two, or three new connections of people entering the category, or people who learned from the last cycle’s mistakes and are finding new niches.
– Nick Pipitone
