Only 4% of institutional investors have not changed their investment approach in light of AI’s impact on software. That single data point from Lazard’s latest Private Equity Secondaries Survey is not just a statistic—it’s a declaration that the old valuation framework for software assets is dead. For those of us who have spent years dissecting whitepapers and auditing token distributions, this feels hauntingly familiar. The same shift is happening in crypto, and the lessons are being written in real time.
Context: The Survey That Speaks to a Paradigm Shift
Lazard’s survey, which reached a broad set of PE secondaries investors, reveals that 91% of respondents now consider “proprietary data plus network effects” as the core moat for software companies. Only 4% said they haven’t changed their methods. The remainder are actively moving capital away from software into other opportunities. This isn’t a casual opinion poll—it’s a capital allocation decision that will reshape the next decade of software investing.
In crypto, we’ve seen a similar pattern. During the 2020 DeFi Summer, I watched projects tout their smart contract code as the moat. But code is copyable. The real moat—on-chain data, user activity, and network density—was always the invisible asset. Now, institutional investors are finally acknowledging that in the AI era, value flows to data, not functions. The same principle applies to blockchain: a protocol’s code is increasingly commoditized, but its data and network effects are irreplaceable.
Core: The Technical Mechanism Behind the Consensus
Why 91%? Because the underlying technology—large language models—has made feature-level software a commodity. A general-purpose LLM can now generate UI, handle basic automation, and even write code. The only thing it cannot easily replicate is a proprietary dataset with a dense user network. This is structurally identical to what happens in Layer 2 networks: the OP Stack and ZK Stack differ in their cryptographic proofs, but the real competitive advantage comes from which stack convinces more projects to deploy chains. Network effects, not technical superiority, determine long-term survival.
In my own experience auditing ICOs in 2017, I identified three token distribution vulnerabilities that could lead to centralization. The projects with the strongest technical whitepapers often had the weakest data moats—they were building features, not ecosystems. Today, the same mistake is being made by projects that obsess over TVL or daily active users without asking: Is the data being generated proprietary? Is the network effect self-reinforcing? The Lazard survey confirms that capital is now flowing to those who can answer “yes” to both.
Contrarian: The Blind Spot in the Groupthink
But 91% agreement is a red flag. When everyone converges on the same narrative, the market is pricing in a future that may not materialize. The survey’s focus on proprietary data and network effects ignores a critical paradox: the more valuable a dataset becomes, the more likely it is to be attacked, regulated, or shared. In crypto, we’ve seen over $2.5 billion lost to cross-chain bridge hacks—yet the industry still depends on them. The “data moat” narrative may be a self-fulfilling prophecy that leads to data hoarding and centralization, exactly the opposite of the open, permissionless ethos that built crypto.
Furthermore, the survey implicitly assumes that AI capabilities will continue to improve at the current pace. But what if the next generation of models (synthetic data, federated learning, long-context windows) can infer the value of “proprietary” data without direct access? The same risk exists in crypto: if cross-chain interoperability matures, the network effect of a single chain becomes less sticky. The moat you thought you had may evaporate.
Takeaway: The Next Narrative in Crypto
“Truth over hype. Always.” The Lazard survey is a wake-up call for crypto investors. The days of valuing a project by its GitHub commits or its gas usage are ending. The new metrics will be about data uniqueness, network density, and AI integration capability. Trust is the only currency that matters, and the projects that accumulate verifiable, proprietary data—while maintaining security—will be the ones that survive the next bear market.
From my experience stabilizing our team during the 2022 crash, I learned that the noise of hype always fades. The signal that remains is the value of real, resistant data assets. Noise filtered. Signal preserved.
So the question every crypto investor should ask themselves: Are we overvaluing network effects at the expense of data integrity? The answer will determine who wins the next cycle.
