The Lazard survey dropped like a depth charge in the private equity world this week. By now, you've seen the headlines: 91% of institutional investors now view proprietary data and network effects as the only real moat in software. Only 4% haven't changed their investment approach. The rest are either waiting or moving capital elsewhere.
But here's the part that keeps me up at night: if the smartest money in traditional finance is running from software that can't prove it owns unique data and a sticky network, what does that mean for crypto? We've spent years telling ourselves that code is the moat—that smart contracts, open-source innovation, and decentralization are our shields. But AI is eating code for breakfast. The Lazard report is a mirror, and it's reflecting a truth we've been too busy building to see: in a world where AI can generate Solidity contracts in seconds, the only moat that matters is the one that exists outside the codebase.
— Root: The 2022 Bear Market
Context: The Decentralization of the Moat
Let's step back. The Lazard survey is a private equity secondary market poll—essentially, a snapshot of what the most sophisticated allocators of capital think about software valuations. The findings are stark:
- 91% believe proprietary data + network effects are the primary moat.
- Only 4% have not changed their investment methodology.
- The majority are either "waiting for clarity" or shifting funds to other asset classes.
For context, when a survey shows 91% agreement on any topic, it's not a signal—it's a consensus shockwave. That level of alignment means the market has already priced in a paradigm shift. The old valuation framework—EV/Revenue multiples based on growth rates, gross margins, and net dollar retention—is dead. The new framework, still forming, will discount software based on AI exposure and premiumize based on data defensibility.
Now, transpose this to crypto. Our industry has always had a complicated relationship with moats. Open source means anyone can fork. Composability means protocols are legos. But we've clung to the idea that first-mover advantage, liquidity depth, and smart contract security are enough. The Lazard survey suggests that in the age of AI, those are necessary but not sufficient. The real moat is something that AI cannot replicate—a closed-loop data network that grows stronger with every user interaction.
Where does crypto stand? We have public blockchains where all data is visible. We have user bases, but they're often scattered across chains. We have liquidity, but it's fragmented. The question is: can a crypto protocol build a proprietary data moat when the ledger is transparent by design? The answer is more nuanced than you'd think.
Core Insight: The Data Paradox of Blockchain
I've been in this space long enough to see the cycles. I co-founded TrustChain in 2017, spent DeFi Summer auditing Uniswap's governance, and weathered the 2022 Bear Market running a resilience hub for junior developers. Through all of it, one pattern repeats: the protocols that survive are not the ones with the best code—they're the ones with the deepest wells of unique data and the strongest network effects.
Consider Uniswap. Its smart contract code is open source. Anyone can fork it. But Uniswap's moat isn't the code—it's the liquidity data. The order book simulation engine, the routing algorithms, the historical price feeds—all of that is proprietary data built from years of network activity. When a new DEX launches, it can copy the code, but it can't copy the data. That's why Uniswap still commands 70% of DEX volume. The code is a commodity; the data is the asset.
Now layer AI on top. Imagine an LLM trained on every on-chain transaction history. It could predict liquidity shifts, optimize routing, and even generate new trading strategies. The protocol that owns that data—and the network that generates it—has a moat that no AI can replicate. Because the AI itself is trained on the data, but the data is generated by the network. It's a flywheel that feeds on itself.
This is the core insight from the Lazard survey applied to crypto: the value anchor is shifting from code to data and network. In practice, that means:
- Layer 2 rollups that optimize for data availability (DA) are overhyped unless they also generate unique on-chain data patterns. 99% of rollups don't generate enough data to need dedicated DA—they're just moving blocks around. The real value is in the transaction data that reveals user behavior.
- DeFi protocols that build on Uniswap V4's hooks will see a complexity spike that scares off 90% of developers, but the ones who master it will own the data layer. Hooks turn DEXs into programmable LEGOs, but the real prize is the data those hooks generate.
- DAOs that rely on delegation will become more centralized—users are too lazy to research and just delegate to KOLs, creating a data oligopoly. The governance data becomes a moat for the few who actually participate.
I saw this firsthand during my DeFi Summer deep dive. We audited Uniswap's governance and published a 50-page white paper on democratizing liquidity. The most valuable part wasn't the code analysis—it was the data on how token holders actually voted. That data showed that 70% of delegates held less than 1% of the voting power, creating a power law distribution. The data was the real story.
— Root: DeFi Summer
Contrarian Angle: The Transparency Trap
Now for the contrarian take. The Lazard survey's emphasis on proprietary data moats assumes that data can be locked away. But in crypto, data is transparent. Anyone can read the Ethereum blockchain. So how can a protocol claim a data moat?
The answer is: contextual data vs. raw data. Raw transaction data is public. But contextual data—the annotated, labeled, and enriched version—is proprietary. Think of it as the difference between a library of books and a library with a trained librarian who knows exactly which book to recommend. The raw data is the books; the contextual data is the librarian.
AI exacerbates this. An LLM trained on raw blockchain data will produce generic insights. But a protocol that builds a vector database of labeled transaction patterns, with user tags and intent signals, creates a data asset that is both unique and valuable. That's the moat.
But here's the trap: the Lazard survey might be overestimating the durability of data moats in a world of synthetic data and federated learning. If AI can generate synthetic transaction data that mimics real user behavior, the uniqueness of proprietary data erodes. We saw this in 2026 when I convened the Autonomous Agent Accountability Charter—we realized that AI agents could generate on-chain data that was indistinguishable from human behavior. The data moat became a minefield of noisy signals.
Furthermore, the survey's 91% consensus might be a self-fulfilling prophecy that ignores the role of trust. In crypto, the social contract matters more than data. The Lazard crowd is private equity—they're used to controlling data. But in crypto, the community is the protocol. Code is law, but people are the protocol. Governance isn't a feature; it's a product. The moat that survives AI is not data—it's trust. And trust is built through transparency, not proprietary data.
— Root: The 2022 Bear Market
Takeaway: The Vision Forward
So where does this leave us? The Lazard survey is a signal, not a verdict. It tells us that the most sophisticated capital allocators are reassessing software value. For crypto, this is a wake-up call to stop pretending that code alone is a moat. The protocols that will thrive in the AI era are those that:
- Generate unique, contextual data through network effects—think Uniswap's routing data, Aave's credit scoring data, or ENS's domain data.
- Build AI-native data pipelines that turn raw on-chain data into actionable intelligence—this is the next trillion-dollar opportunity.
- Maintain community trust through transparent governance and user-centric design—because when AI can replicate your code, the only thing left is human connection.
I've seen this movie before. In 2022, when the bear market hit, everyone panicked. But the protocols that survived—the ones that doubled down on community, education, and data—emerged stronger. We built the Resilience Hub, mentored 200 junior developers, and created a repository of 300+ resources. That wasn't code; it was trust. And trust is the only moat that AI cannot steal.
The future of crypto is not in writing smarter contracts. It's in generating smarter data. The Lazard survey is a mirror, and it's saying: "Your code is not your castle. Your data and your people are." Listen to it.
— Root: DeFi Summer