Chapter 25 · Understand a Crypto Project in 30 Minutes
What should you inspect first in an unfamiliar project?
- Capabilities practiced
- ResearchSystem Thinking
- Hands on
- Take a project you saw today, fill all thirteen cells in 30 minutes, and mark any unanswered ones.
- AI Lab
- Have AI fill in the thirteen fields first. Your tasks: verify the source of every field and cross out unsupported content.
A Real-World Problem
A colleague shares a link on Monday: take a look at this.
The homepage says next-generation modular liquidity layer, with icons, animation, and a slogan. It promises a launch, 12 supported chains, and 180,000 community members. Twenty minutes of documentation leave you with unfamiliar terms.
Then comes what does it actually do?
You cannot answer. You were reading material rather than seeking answers.
Dozens of polished projects appear daily. You need a defensible assessment within half an hour, or an honest account of what remains unknown.
This chapter asks what to inspect, in which order, and when to stop.
It assembles the previous 24 mechanism chapters into a portable format, supporting the rest of Stage 5.
Thought Experiment
Leave blockchains briefly.
Assess whether a shop on an unfamiliar street can survive next year, with half an hour and no financial statements.
What would you ask?
Most can list:
- What does it sell, and whose problem does it solve?
- Who visits, incidentally or deliberately?
- Who pays, and how much remains with the shop?
- Who funds rent, and for how long?
- Which competitors exist, and why choose it?
- What have its founders done before?
- What changes will arrive in three months?
- What is the likeliest failure cause?
Questions stay fixed while materials change. Investigate with the framework already in hand.
Suppose staff decline to explain who pays rent.
How do you record it?
Two approaches:
A: The owner probably pays, since it is their shop.
B: Rent payer: asked, no answer. Leave blank.A fills the table and invalidates the judgment, turning guesses into remembered facts.
B preserves evidence: the shop chooses not to disclose something, which may itself matter more than an answer.
Crypto adds more questions about tokens, rules, and permissions, plus answers you can verify on-chain yourself.
Your Decision
How do you use the half hour?
What Happens
The approaches agree:
Question organization is the bottleneck, not information volume.
Materials exceed available time. Scarcity lies in fixed, distinct, verifiable questions and a stopping rule.
Each method has a role:
| Method | Answers | Cannot answer | Role |
|---|---|---|---|
| Official documents | Claimed purpose | Truth and omissions | Main source material |
| Prices and community | Attention | Underlying substance | One cell |
| On-chain records | Actual activity | Its reason | Verification |
| Fixed table | Missing coverage | Content itself | Framework |
Use the table as structure, documents as content, chains as verification, and attention as one input.
Sources differ in evidential strength:
| Level | Meaning | Confidence | Label |
|---|---|---|---|
| Primary on-chain facts | Contracts, holders, transactions, authority | Reproducible | Verification location |
| Official claims | Documents and announcements | Incentives shape framing | Official statement |
| Third-party summaries | Platforms, research, media | Definitions may be unclear | Source and scope |
| Community and rumors | Chats and anonymous claims | Lowest | Unverified |
| Your inference | Reasoned from evidence | Hypothesis | Inference |
Unlabeled inference is a common costly mistake. Later you remember the table, not which cells were guessed.
Build a Model
Thirteen cells
The shared format sits in the Analysis Toolkit:
Whose problem is it solving, and what exactly is the problem?
Who is actually using it today?
What are the protocol-layer rules? Who can change them?
What does the user actually see and do?
What sector is it in, how large is the market, and where are we in the cycle?
Is there a token? What are its supply, unlocks, and uses?
Where does protocol revenue come from, and who receives it?
Who is doing it, what have they done, and can it be verified?
Who invested, at what price, and when does it unlock?
Is the community made of users or airdrop hunters?
Who else is doing the same thing, and why will this one win?
What will change its position in the next three months?
What single event is most likely to reduce its value to zero?
Three points.
First, a coordinate system whose cells are distinct and expose key structure, rather than exhaustive project trivia.
Second, six cells describe what it is, four how it survives, and three what happens next.
| Group | Cells | Question |
|---|---|---|
| What it is | Problem, User, Protocol, Product, Market, Token | Activity and structure |
| How it survives | Revenue, Team, Capital, Community | Money and people |
| What comes next | Competition, Catalyst, Risk | Three months to a year |
Third, empty cells are risks.
Three possible reasons:
| Reason | Meaning | Record |
|---|---|---|
| Available but not found | Research incomplete | Pending investigation |
| Absent from materials | Project chooses silence | Undisclosed |
| Project cannot explain | No established answer | Unclear |
Do not substitute guesses. The third deserves special scrutiny: no payer may be the answer itself.
Allocate 30 minutes
Time is a constraint. Longer reading can replace judgment with familiarity.
- Problem and User first
- Revenue and Token next
- Fill the remaining nine
- Risk and one judgment last
Suggested allocation:
| Time | Task | Cells |
|---|---|---|
| 0–5 minutes | Whose task becomes easier? | Problem, User |
| 5–12 minutes | Payers, recipients, token purpose | Revenue, Token |
| 12–22 minutes | Remaining content and verification | Other nine |
| 22–27 minutes | Label sources and remove unsupported claims | All |
| 27–30 minutes | Risk, conclusion, follow-up list | Risk |
Source labeling separates reading from assumption while memory is fresh.
Three red flags
Record these directly:
- Problem requires jargon. Preserve the original and mark clarification needed if plain language fails.
- Revenue says token incentives. That is expense. Apply Chapter 17's payer questions and Chapter 26's calculations.
- Token is more detailed than Product. Use Chapter 20's flow model and Chapter 19's unlock schedule while recognizing the emphasis.
From cells to judgment
The table supplies evidence, not the conclusion.
Write three sentences:
Purpose: It makes ___ easier / cheaper / possible for ___.
Risk: The likeliest failure cause is ___.
Confidence: ___ cells verified on-chain; ___ unavailable.More than five unavailable cells should prevent a firm judgment.
What It Is Called
Thirteen fixed questions describing a project.
A checklist revealing omissions; use every cell.
Twenty completed canvases enable direct comparison, making missing revenue obvious across peers.
Whose task becomes easier, cheaper, or newly possible?
Explain without jargon. Modular liquidity abstraction is vague; using one deposit as collateral across three chains is concrete.
An empty Problem undermines the other cells, echoing Chapter 1's balance question.
How protocol income connects to holders.
The Revenue-to-Token link is often missing despite genuine fees.
Following Chapter 20, ask what holders gain from one additional revenue dollar. Mark absent connections unestablished.
Confirmed events within three months that change circumstances.
Confirmed dates include launches, unlocks, votes, and approval deadlines. Vague plans do not qualify.
This cell adds time to description. Chapter 19's unlock steps are catalysts.
The likeliest single failure cause.
Use Chapter 24's categories to prioritize, rather than copy the full list.
That item becomes the first check on future reviews.
A valid cell entry, equal in status to factual answers.
Distinguish pending, undisclosed, and unclear.
Honest missing entries determine quality, following Chapter 24.
Hands-on
Choose a project first seen today, so familiarity cannot let you skip the process.
Set a timer and stop at 30 minutes. Empty cells are an output, not a failure.
0–5 minutes: find one sentence.
Read the homepage and documentation introduction, seeking whose task becomes easier.
Rewrite in your own words. If impossible, copy the claim and label it not explainable in plain language.
Problem and User must contain something, even unclear.
5–12 minutes: trace money.
Find payer, recipient, and token existence and purpose.
Search fees, revenue, treasury, token uses, and tokenomics. Record specific source sections, not only conclusions.
Clear rates with missing destinations are common. Record the destination as undisclosed.
12–22 minutes: fill and verify.
Four cells support direct checks:
| Cell | On-chain check |
|---|---|
| Protocol | Address, proxy status, upgrade authority |
| Token | Supply, holders, top-ten share |
| Community | Recent interactions versus advertised members |
| Capital | Public investor addresses and unlocked proportions |
Verify addresses from official documents in explorers, applying Chapter 23.
22–27 minutes: label sources and remove unsupported content.
Assign each cell a level:
[On-chain] Verified myself at …
[Official] Document section X / announcement …
[Third party] Platform / report, definition …
[Unverified] Community or rumor
[Inference] Derived from other cellsReplace unsupported inferences with not found. A less polished table can be more useful.
27–30 minutes: conclusions and follow-up.
Purpose: It makes ___ more ___ for ___.
Risk: Most likely failure is ___.
Confidence: ___ cells verified on-chain; ___ unavailable.
Follow-up: 1. … 2. … 3. …A follow-up list makes the canvas reusable, rather than disposable.
Completed example
A fictional but internally consistent project, Northwind, finances short-term real-world receivables using stablecoin deposits.
| Cell | Content | Source |
|---|---|---|
| Problem | Traders wait 60–90 days for payment; stablecoin holders have idle dollars | Official section 1 |
| User | Individuals and small institutions supply funds; partner-introduced receivable issuers borrow | Official; issuer list undisclosed |
| Protocol | Pool contract plus off-chain review, with multisig-adjustable parameters | Contracts and multisig verified |
| Product | Deposit stablecoins for receipts and request redemption through a web app | Official and inspected interface |
| Market | On-chain real assets, at least six peers | Third-party category page |
| Token | Governance token not yet trading; supply and allocations published, vesting undisclosed | Official and missing information |
| Revenue | 0.5% annual management fee plus issuer transaction fees; destinations and control undisclosed | Official rates, missing ownership |
| Team | Three named founders; two histories verifiable, one unavailable | Official, partially verified |
| Capital | Two rounds and four investor names; amounts and prices undisclosed | Official and missing information |
| Community | Claims 40,000 members; three-digit recent contract-interaction count | Official versus on-chain |
| Competition | Six peers differ in bad-debt responsibility; Northwind disclaims it | Third-party and official |
| Catalyst | Token launch and regulatory-response deadline within three months | Official announcements |
| Risk | Off-chain issuer default without on-chain recovery | Inference from Protocol and Competition |
Three sentences:
Purpose: Stablecoin holders lend to trading companies for receivable discount income.
Risk: Off-chain issuers default, with no on-chain recovery and no protocol backstop.
Confidence: Three cells checked on-chain, four undisclosed, one inferred.
Follow-up: Fee destination, vesting schedule, and bad-debt loss order.Four undisclosed cells concentrate around money and control, itself informative. Risk is labeled inference with supporting reasoning. Inference is permissible when identified.
AI Lab
Complete the manual Lab first to establish a basis for checking.
Models are fast and useful here, but can fill blanks too confidently. Compression is easier for them than acknowledging missing answers.
Organize [project name and official URL] into:
Problem / User / Protocol / Product / Market / Token /
Revenue / Team / Capital / Community / Competition / Catalyst / Risk
For every cell provide:
1. Content; Problem must use a jargon-free sentence
2. Specific source section, page, or address; otherwise "no source"
3. Level: on-chain / official / third party / unverified / inference
Additional rules:
- Separate payer from final revenue authority.
- Identify promotional Community counts without on-chain support.
- Catalysts need confirmed dates or deadlines.
- Risk names one likely failure with reasoning.
Missing answers: pending / undisclosed / unclear.
Never invent plausible content to complete the table.
Timestamp every number.Remove no-source cells first, then verify remaining evidence.
Frequent errors:
- Slogans masquerading as Problem. Test whether a non-crypto reader can explain it.
- Incentives counted as Revenue, addressed in Chapter 26.
- Rumors counted as Catalyst, without dates or verification.
- Generic Risk descriptions, which Chapter 24 showed convey little.
Compare your omissions, its omissions, and disagreements in shared cells. The third produces the most learning, and Chapter 31 formalizes this comparison.
Verify these points yourself after using AI
- Does every cell cite evidence? Remove unsupported content even if plausible.
- Have contract and token addresses been opened in explorers?
- Does Problem use plain language rather than a rewritten slogan?
- Does Revenue distinguish earnings from token incentives?
- Does Community use promotion counts or verified activity?
- Are catalysts confirmed rather than rumored?
- Does Risk name one prioritized issue rather than a generic list?
- How many cells say not found? None suggests invented completion.
Real-World Case
Documentation clearly states fees, LP shares, and protocol shares.
But the protocol's destination address and authority are missing, with unanswered governance questions.
No conspiracy is needed; incomplete documentation produces the same research gap.
That gap propagates into unconfirmed token capture and treasury-control risk.
A project counts social-platform memberships.
Chapter 23's method yields three-digit recent interacting addresses.
Awareness and usage are different measures.
Record both: the two-order-of-magnitude gap is informative.
A detailed report replaces confirmed catalysts with vague ecosystem plans.
Later a large scheduled unlock moves the market. The public date was omitted from the canvas.
Chapter 19's unlock steps are catalysts. A brief schedule check can prevent costly omissions.
A polished report leaves no blanks.
Verification reveals expectations and anonymous claims mixed with established facts.
Failure to label inference removes readers' ability to judge confidence.
Source-labeling time is essential. An incomplete transparent canvas can be more dependable than a full opaque report.
Change One Variable
More detail becomes available, with opposing effects.
Benefit: Protocol, Capital, and Competition receive deeper checks.
Cost: familiarity and sunk effort make unsupported acceptance easier.
Separate timed intake from deeper investigation. First record judgments and gaps, then research the follow-up list.
Product, Revenue, empirical Community and Competition, and on-chain checks may not exist yet.
Early projects can legitimately lack these, but remaining cells bear more evidential weight, especially Problem, Team, and Token.
Detailed economics without a product reveals priorities. Chapter 30 explores the pattern.
This is an answer, not missing information.
First: temporary fee switch or no charging mechanism? Contracts distinguish them.
Second: capital and treasury become the operating lifeline. Inspect holdings and runway, as Chapter 26 does.
The main risk may be funding exhaustion rather than attack.
The pattern is already a conclusion.
Money-and-team gaps suggest nondisclosure; protocol-and-product gaps may suggest an unbuilt project.
Write not assessable within available evidence.
That is a concrete, justified finding, revisable when disclosure improves.
Takeaway Questions
Problem must use jargon-free language.
A non-crypto friend should be able to relay your sentence. Otherwise you retained the project's claim rather than answering.
Replace the chain with a company database: what breaks?
Custody or verifiability may explain necessity. If nothing breaks, label necessity unestablished and carry it into Risk.
Prioritize Risk through who loses first.
Without an order, structure remains unclear. Use Chapter 24's categories to prioritize and trace loss bearers.
AI reduces material-structuring effort rather than verification effort.
Let it organize; you remove unsupported content. Chapter 31 expands this into a pipeline.
Chapter Check
Three causes differ:
- Pending: your search is incomplete.
- Undisclosed: deliberate absence. A finding.
- Unclear: no answer exists. The strongest concern.
Silence about revenue control can itself describe the project.
Guesses erase that information and create an indefensible table.
The Problem-to-Revenue connection.
Who benefits, and who pays? Connecting them establishes the basic business structure.
If they differ, use Chapter 17's payer motivation and duration questions.
Protocols paying users through incentives require Chapters 26 and 28's sustainability tests.
Five evidence types have very different strength.
The table survives in memory while source distinctions fade.
Without labels, guesses become facts, confidence adjustments disappear, and replication fails.
Labeling exposes unsupported cells while memories are fresh.
Revenue and Risk.
Revenue often confuses incentives or total fees with protocol earnings. Separate payers and authority.
Risk often produces a generic paragraph. Require one prioritized issue and reasoning.
Also count missing answers. None suggests artificial completion.
Check:
- Was it strictly timed?
- Plain-language Problem?
- Payer and authority separated?
- Source labeling and unsupported removals?
- Confirmed catalyst dates?
- One Risk?
- Missing cells? None suggests guessing.
Three to five unknowns, often around money and control, are normal. Honest incompleteness is expected.
After four weeks, compare canvases. Shared gaps can reveal structural sector problems.
Key idea
Fill all thirteen Project Canvas cells before judging; any cell you cannot fill is a risk.