In August 2020, I put $15,000 AUD — every dollar I had — into a yield farm I had “researched” for about forty minutes. My diligence spreadsheet had nine columns. I filled in three. The other six said, effectively, nothing: no audit, no identifiable team, no unlock schedule, no contract owner, no revenue, no reason. I looked at those six empty boxes, and instead of closing the laptop, I wrote myself a story about why the emptiness was fine.
The story took forty minutes to write too. It just happened to cost $15,000.
I've thought about those six boxes for five years. I reverse-engineered the exploit that drained them, published a repo, wrote the post-mortem, and rebuilt my instincts around a single rule I now repeat to every founder who asks me for feedback: an empty cell is not a neutral fact. It's a claim you haven't heard yet.
So when a colleague forwarded me a research template last week that had been run, end to end, against a source document that turned out to be nothing — every dimension, technology, tokenomics, market structure, ecosystem position, regulation, team, risk, narrative, supply-chain transmission, each one returning the same three letters — I felt something I didn't expect. Not frustration. Not the eye-roll I reserve for lazy analyst output.
Relief.
N/A. Not one invented number. Not one confident sentence built on air. The analyst who produced it refused to guess, and — this is the part that matters — refused to dress the refusal up as a finding. “Cannot be assessed” is not the same as “no risk.” They wrote that down explicitly, so nobody downstream could misread a blank as a green light.
That document is the most honest thing I've read in crypto this year. It is also, quietly, a mirror.
Here's why. Truth in blockchain isn't a vibe, and it was never a marketing category. It is the specific, unglamorous property that the state of a system is publicly readable and independently recomputable. You don't have to take my word on a balance. You can check it. That's the whole innovation. Everything else — the tokens, the conferences, the podcasts, the liquidity — is downstream of that one boring guarantee.
And yet the industry that invented verifiability runs almost entirely on unverifiable narrative. We didn't build a trustless system so we could trust strangers' slide decks instead. We didn't spend a decade arguing about consensus mechanisms so that diligence could be replaced by a feeling in a Telegram group.
Which brings me back to the blank template, and why I think it's a technical document rather than a bureaucratic one.
Take the first dimension: technology. “N/A” there usually doesn't mean the technology is unknowable. It means nobody looked. Whether a protocol's contracts are upgradeable is a five-minute check — is there a proxy admin, who holds those keys, is the implementation slot mutable. Whether an audit exists is a search. Whether deployed bytecode matches the audited commit is verifiable. When a research output returns N/A on technology, the honest translation is: the person producing it did not have access, or did not have time, or did not have permission. Emptiness here describes the researcher, not the protocol.
Tokenomics is the dimension that should make you angry, because it's where emptiness is most often deliberate. Vesting schedules live on-chain. Cliff dates are readable. What is frequently missing is not the data — it's the labeling of the data. A wallet holding 22% of a supply, unlabeled, isn't unknown. It's unattributed. Those are different problems with different fixes, and conflating them lets a project look mysterious when it is merely undisclosed.
Ecosystem position has the same shape. Developer counts, deployment activity, retention curves — all measurable, all frequently left as N/A because the measurement is tedious rather than impossible. When a report tells you a protocol's ecosystem is “growing,” ask for the commit graph.
Then there's governance, where I've spent most of my professional life.
“Code is law” has never survived contact with an upgrade key. In practice, what governs a DAO is not the voting contract. It's whoever can call the proxy admin function at 3 a.m. That's almost always a multi-sig with five to nine signers, and those signers are usually the same four funds and two core developers you'd find on the cap table. A governance forum will debate a parameter change for three weeks at 4% turnout, and the multi-sig will execute it in ninety seconds. Neither fact is a scandal. Both are facts. When a governance dimension comes back N/A, what you've actually learned is that nobody mapped the admin key — which is the single most load-bearing fact about the system.
Layer 2 is its own case study in permanent N/A. Decentralized sequencing has been a PowerPoint slide for two years. I've been asking the same question since 2022: who orders the transactions, and what happens if that operator goes down or decides to reorder? Today the answer, for most of the networks holding the most deposits, is still one operator running one node. That's defensible as an engineering tradeoff. It is not defensible as a marketing claim. But try finding the sentence “single sequencer, no fallback, no timeline” on a landing page. The data is public. The framing is optional. Emptiness, again, means unexamined, not unknowable.
Stablecoins taught me the same lesson from the opposite direction. For years, Western analysts wrote about adoption in emerging markets as an ideological story — freedom money, censorship resistance, the unbanked. Then you actually go look, and the driver is simpler and sadder: local currency inflation, capital controls, remittance costs, and a family trying to keep a month of groceries from evaporating. The ideology was the narrative. Inflation was the data. Both were sitting right there. We just picked the one that made a better headline.
Which brings me to the uncomfortable part.
The blank template is honest. Honest is not the same as sufficient.
I've watched enough research desks to know how this goes. Nobody fabricates on purpose. What happens is subtler: an empty cell creates social discomfort, discomfort creates pressure, and pressure produces a sentence like “the team is well-positioned for the next phase of growth.” That sentence isn't a lie. It's also not information. It's filler occupying the exact slot where a verifiable claim should sit, and it makes the document look complete. An analyst who refuses to write it looks lazy. An analyst who writes it looks employed.
That's the incentive. In a bull market, the filled-in deck gets funded and the blank one gets ignored. We didn't design a market that rewards accuracy. We designed one that rewards the appearance of coverage — and then we act surprised when coverage is all we get.
So the discipline I want to argue for is narrower than “be honest.” Refusing to guess is necessary but not sufficient, because a research process that returns N/A nine times out of nine has stopped researching. A blank is a to-do list, not a verdict. The right output isn't a finished document with empty cells sitting inside it. It's a document that says, in order: here is what is verifiable, and here is the block explorer link; here is what is unverifiable, and here is who would have to tell me; here is what I could not obtain, and here is the next person I'm going to ask.
That's a different artifact. It's less satisfying to read. It's much harder to fund. It's also the only version of this work that compounds, because every blank you eventually fill is a real asset, and every confident sentence you invent is a liability with a delay on it.
The next eighteen months will be loud. There's institutional money on the tape, an ETF-shaped narrative, a regulatory mood that has finally stopped being the only story, and a fresh wave of projects raising on decks that would return a lot of N/A if anyone ran the template. Most of that capital will be deployed by people who cannot read a vesting contract and will not ask who holds the upgrade key.
That isn't a reason to despair. It's a reason to be specific. The edge in the next cycle won't be finding the project nobody has heard of. It'll be being the person in the room who asks who the sequencer is, who signs the multi-sig, and when the next cliff unlocks — and who can answer with a link instead of an adjective.
I still keep the 2020 spreadsheet. I didn't lose fifteen thousand dollars. I bought six empty boxes, and they've been the most expensive and most useful research tool I own.
The blanks are still there. I just fill them differently now.
