The most expensive mistake I almost made with Propex never cost me a dollar. It almost cost me the whole company.

Early on, I wanted Propex to be the marketplace. The Amazon of tokenized real estate. One brand, one storefront, every property in Bali, Barcelona and Phuket listed under my logo, millions of buyers showing up to browse. It's the obvious dream. It's also the dream that kills most tokenization startups before they ever onboard a single real asset.

I didn't build that. I built the rails underneath it instead. Here's why, and what it actually cost me to make that call.

Every tokenization founder I meet starts in the same place. They want to be the destination. The platform people go to. So they spend their first year building a consumer marketplace, polishing the buy button, obsessing over the checkout flow.

Then they go looking for inventory. And the inventory isn't there.

Because quality supply doesn't come from a website. It comes from operators — developers, agencies, property managers, the people who already control real buildings, real tenants, real cash flow, and real trust in their local market. Without them, your beautiful marketplace is an empty mall. You're standing in the food court at 9am with the lights on and nobody coming.

I watched enough of those empty malls get built that I made a different bet. The bottleneck in tokenized real estate was never demand. Crypto liquidity has been sitting in wallets for years looking for something real to do — including real yield. The bottleneck is credible supply. And credible supply lives with operators, not with me.

So the question stopped being "how do I get buyers?" It became "how do I make it stupidly easy for the people who already have the assets to bring them on-chain?"

Propex became infrastructure. The pipes. White-label storefronts that operators run under their own brand, on their own domain, with their own buyers. The legal templates, the smart contracts, the ownership logic, the distribution rails — that's the part I own. The customer relationship stays with the operator.

But the storefront was never the point. What matters is what flows through the pipe. Propex captures the money flow of real estate — the rent, the appreciation — and puts it on-chain, where it becomes real yield served to capital that's already sitting there. The building stays where it is. The token carries the ownership rights. The smart contract distributes the rent automatically.

The first operator on Propex is BalirealtX. They don't sell "Propex tokens." They run their own branded marketplace for Bali property, and Propex is the engine in the background. The model underneath is deliberately boring: an LLC wrapper plus a third-party manager handles the rentals and sends the cash flow to token holders, and the only performance number anyone publishes is existing results, never a promise of future yield. We list property that already earns. In Bali right now, income-ready assets carry a 20%-plus premium over off-plan (REID, Q1 2026) — because real, existing yield is worth more than a pitch deck.

I'm not a lawyer, and none of this is legal advice. But that structure isn't an accident. It's built so the return comes from owning a SPV and a real operator doing real work — not from me promising to make you rich. That's the difference between infrastructure you can trust and the "sold out in 45 seconds, 25% yield" pitch that gives this whole industry a bad name.

Being the rails instead of the store meant giving up the thing every founder's ego wants: the logo on the front door. Buyers will never know my name. That was the price. What I bought with it was the ability to add supply through a network that already has it, instead of begging for it one business at a time. Every operator we onboard isn't just another storefront. It's a new stream of real yield flowing onto the same rail.

This used to look like a contrarian bet. It isn't anymore. a16z put it plainly this year: the best businesses sit inside the flow of value and take a clip. The money flow is the moat.

That's the whole reason I'd rather own the pipe than the storefront. Storefronts are visible, replaceable, and easy to copy. The flow of real income running through them is not. Finance itself is commoditizing — anyone can spin up an app, a card, a yield-bearing balance now. Deel pays around 4.5% on payroll balances. Whop pays 4% on creator revenue. The account stopped being the moat. The supply producing the yield became it.

In real estate, that supply is brutally hard to assemble — a four-trillion-dollar asset class fragmented by a thousand rulebooks, different in every jurisdiction. Which is exactly why a flexible, neutral rail wins it, and why it gets more valuable as more operators plug in. More supply, more flow, more participants, more supply. a16z again: money flow plus network effects is one of the most durable structures ever built.

I'm not predicting a boom. I just refuse to compete for the part of the market that's about to get crowded, and built the part that gets more valuable as everyone else piles in.

The lesson isn't "build infrastructure." Infrastructure is the right answer for me, in this market, at this moment. It might be wrong for you.

The real lesson is older and more annoying than that: pick one. The most dangerous place to stand as a founder is in the middle, half storefront and half infrastructure, doing both badly because you couldn't bear to give up either. Every company that tries to be the marketplace and the rail at the same time ends up too consumer-facing to be trusted as neutral infrastructure, and too infrastructural to win the consumer.

Choosing what not to be is the hardest call you make early, and the one nobody claps for. Nobody congratulates you for the marketplace you didn't build. But that decision shapes everything that comes after it.

I gave up being the front door so I could be the foundation — the rail that brings credible supply on-chain and turns real rent into real yield. In a market filling up with front doors, that's the trade I'd make again.

The Nifty Founder is a 2x/week newsletter about building, structuring, and owning in the new economy.