Four Years In
What I got wrong, what we've built, and what's next
Four years ago, I launched a real estate investment banking boutique. The thesis was simple: there are hundreds of innovative real estate operating companies building in niche asset classes—data centers, outdoor hospitality, senior living, cold storage—and most of them are too small or too unconventional for the big banks to care about. Someone should help them raise capital.
The business model was the same one every capital raiser runs:
Get a client;
Use their deal as bait to get meetings with investors;
Learn what those investors want;
Use that knowledge to get the next client;
Repeat.
The product, really, was a Rolodex—and the only way to build it was through a steady stream of dinners, conferences, and cold emails.
It worked, slowly.
And five years in, I can look back at a series of assumptions I held on day one that turned out to be wrong — or at least incomplete. Each time I updated one of those assumptions, we ended up building something new.
What started as a traditional advisory business has become something I didn’t plan for and am still figuring out: a platform that combines media, data, AI, and live events to connect real estate operators with capital.
I want to walk through what I got wrong, what we built as a result, and where it’s going.
“Capital raising is a relationship business.”
This is the thing everyone says. And it’s true—but it’s an incomplete truth that leads to a bad strategy.
Yes, an investor is more likely to take a call from someone they know and trust. But the traditional version of “relationship-driven capital raising” means a senior person manually working a network of a few hundred contacts, built over decades, maintained through dinners and conferences. The knowledge about what those investors want lives in that person’s head. When they leave, the Rolodex leaves with them.
I spent my first two years doing exactly this: flying to conferences; hosting dinners; and getting introduced to one LP at a time. And it struck me that the bottleneck wasn’t the quality of our deals or the caliber of the operators we worked with. The bottleneck was that I could only build relationships at the speed of one human being’s calendar.
The assumption I should have started with: capital raising is a trust business, and relationships are just one way to build trust. Content is another.
That realization led us to start using Thesis Driven—a media company my partner Brad Hargreaves had already built into one of the most-read real estate publications in the country—as a capital markets tool.
We started writing about the asset classes and themes our operators were building in.
We published deep dives on niche sectors, interviewed operators on camera.
We hosted virtual events where investors could hear directly from founders.
The investors came to us. Not because we pitched them, but because they’d been reading us for months and already trusted the perspective. Media turned out to be the best investor acquisition channel in real estate—and almost nobody is using it that way.
“Data is a back-office function.”
When we started taking investor meetings—dozens, then hundreds—I kept a spreadsheet:
what asset classes each investor liked…
check sizes…
preferred structures…
geographic focus…
notes from every call.
It was useful but messy. And I realized pretty quickly that the intelligence we were collecting was more valuable than I’d understood. Every investor conversation was a data point: what’s getting allocated to, what’s out of favor, where the capital is flowing. Individually, each call is just a call. In aggregate, it’s a live map of private real estate capital markets.
So we built a platform to capture and structure all of it. Every investor conversation gets logged. Mandates, preferences, deal history, and relationship maps are all tracked. We now have structured intelligence on thousands of institutional investors, family offices, and RIAs — not scraped from the internet, but gathered from real conversations over years.
What I got wrong was thinking of this as an internal tool. The data layer turned out to be the connective tissue of the entire business—the thing that makes every other piece work better. It powers our advisory engagements, informs our media coverage, and increasingly, it’s a product in its own right for operators who want to understand the investor landscape before they start raising in CapitalStack.
“AI will just make outreach faster.”
When AI tools started getting good enough to draft real prose, the obvious application was email. Build a list, generate personalized outreach, send at scale. And that’s what most people in capital raising are doing with AI right now — producing faster, more efficient spam.
I almost made this mistake.
The instinct is to take the most tedious part of the job (writing 200 investor emails) and automate it. But automating a bad process just produces bad output faster.
The real unlock wasn’t speed; it was matching.
We’ve now brought in a machine learning engineer to build workflows that sit on top of our investor dataset. The system doesn’t just draft emails—it starts upstream.
Which investors in our universe have a mandate that fits this deal?
Which ones have done something similar in the last 18 months?
Who in our network has a warm path to the decision-maker?
How should the pitch be framed differently for a pension fund allocator versus a family office principal?
By the time a message gets written, the AI has already done the work that used to take a senior person days: prospect research, relationship mapping, pitch calibration.
The senior leader stays in control of the message and the relationship, but the grunt work (the thing that made in-house investor outreach impractical for most operators) disappears.
Done right, AI doesn’t make outreach faster. It makes outreach smarter—and keeps the senior leader in the driver’s seat. Done wrong, it’s just a more efficient way to erode the trust you spent years building. The difference is whether you start with data and judgment or start with a template and a send button.
“You can do all of this from behind a screen.”
This is the one I keep relearning.
Media scales… data compounds… AI workflows collapse manual labor…. but, none of it replaces what happens when you put an operator and an investor in the same room.
So we’re building out a live events program—starting with a Capital Markets Summit this summer—designed to do exactly that. Not a generic industry conference with 2,000 people and a trade show floor, but a curated room where the operators we work with meet the investors we’ve identified as genuine fits, with enough shared context that the conversations skip the preamble and get to substance.
The events are the part of the business that doesn’t scale. But they’re where trust gets built in a way that no email, article, or dataset can replicate. And they’re the reason the rest of the stack exists — everything else is designed to make that room as productive as possible.
The stack
When I zoom out, what we’ve built is four layers that feed each other:
Media creates the top of the funnel—operators and investors find us through the writing, the events, and the research. It builds trust at scale in a way that cold outreach never will.
Data captures the intelligence from every conversation and structures it into a living map of who’s investing in what, at what size, and through what structures.
AI workflows sit on top of the data and collapse the manual work of investor discovery, matching, and outreach.
Live events bring it all together in a room, with context, where real relationships start.
Each layer makes the others more valuable.
The media drives the conversations that feed the data →
The data powers the AI workflows →
The AI workflows make the events more targeted →
The events generate the relationships that become the next round of intelligence.
I certainly didn’t plan this. Five years ago, I thought I was building an advisory business with a Rolodex.
What we’re actually building is an ecosystem designed to connect the operators innovating across the built world with the capital to do it—using tools that didn’t exist when we started.
What’s next
We’re still early. The AI workflows are in pilot; the events program is just getting started; and the data platform is good but not where it needs to be. I share all of this not because we’ve figured it out, but because I think the model is right—and I want to build it in the open.
If you’re an operator raising capital in a niche or innovative corner of real estate, or an investor trying to find them, this is what we’re working on.
We’re grateful to everyone who’s been part of the journey so far—the operators who trusted us early, the investors who took our calls, and the readers who’ve been following along. You’ve shaped every part of this more than you know.
More to come,