Step 1 — Define your direction
Mission
What long-term transformation of the system do you believe must happen?
A mission is not a product description — it is a direction of change. It names the large system you believe must be transformed, and the future state that should exist if you succeed. Without a clear mission, founders optimize for product features instead of systemic impact.
Guiding questions
- What large system do you want to fundamentally change?
- What future state should exist if you succeed?
- Why does this transformation matter — who benefits and how?
Example
Accelerate the global transition to sustainable energy by making clean power cheaper and more accessible than fossil fuels.
mission frames system
Step 2 — Map the system
System / Market Context
Describe the broader system where value is created or destroyed.
Every startup operates within an existing system of actors, flows, and incentives. Before finding a constraint, you need to understand the system well enough to see where value is created, where it is captured, and where it is lost. The system is the environment your company will try to change.
Guiding questions
- What system or market are you analyzing?
- Who are the key participants and what do they each want?
- How does value currently flow between them?
- What determines winners today?
Industry / Domain
Key Actors
Value Flow
system reveals constraint
Step 3 — Identify the constraint
Structural Constraint
The dominant friction preventing the system from scaling. One statement.
A structural constraint is not a surface complaint — it is the mechanism that prevents the system from functioning better. Great companies don't solve complaints; they remove the structural bottleneck that generates complaints. One well-defined constraint is more valuable than ten symptoms.
Guiding questions
- What repeatedly breaks or stalls in this system?
- Where is the biggest structural inefficiency?
- What do participants consistently complain about?
- What would need to be true for the system to 10x in value?
Example constraint
"Accommodation supply exists but is inaccessible to travelers — matching is fragmented, trust is absent, and discovery is broken."
Weak constraint
"People want better apps." — This is a preference, not a structural limit. It explains nothing about why the system can't improve.
Strong constraint
"Restaurants cannot coordinate delivery logistics without a dedicated driver network, which no single restaurant can afford to build." — Structural, specific, explains the mechanism.
constraint validated by evidence
Step 4 — Validate with evidence
Evidence / Signals
What real-world observations prove this constraint exists?
Evidence prevents the canvas from producing purely speculative thinking. Before designing a solution, verify that the constraint is real and persistent. Look for workarounds (people solving the problem imperfectly), complaints (forums, reviews, repeated friction), and failed solutions (products that partially addressed it but couldn't scale).
Guiding questions
- What signals show this problem is real and persistent?
- What data or observed behaviors support your constraint claim?
- What workarounds do people use today — and why are those workarounds inadequate?
Example signals
Users already attempt this with spreadsheets. Forum threads discuss this problem repeatedly with no satisfactory solution. Existing products partially solve it but retain fundamental limitations.
People use workarounds (spreadsheets, manual processes, hacks)
Forums and reviews show repeated, unsolved complaints
Prior products attempted this and failed or stalled
Data shows persistent inefficiency that hasn't improved
evidence suggests kernel
★ This is the kernel — the structural mechanism that unlocks the system.
Step 5 — Find the kernel
★Kernel Insight
The fundamental mechanism that removes the constraint. True. Widely disbelieved by incumbents.
The kernel is the structural mechanism that removes the constraint. It is not a feature, a product, or a market observation. It is the insight that makes a new category of value possible. A good kernel is true, widely disbelieved by incumbents, and generates a compounding system when acted on.
Guiding questions
- What structural change could remove the constraint entirely?
- What mechanism unlocks new value that currently doesn't exist?
- Why do incumbents believe this is impossible or irrelevant?
- Can you state the insight in one falsifiable sentence?
Example kernel
"Trust infrastructure — verified reviews and payment protection — unlocks private homes as scalable accommodation supply, without the need for hotel infrastructure."
Not a kernel
"We will build a better CRM with AI features." — This is a product description. It names a solution, not a mechanism. Incumbents aren't threatened.
Kernel insight
"If sales follow-up is written automatically at the moment of intent, SMB conversion rates triple — because the bottleneck is composition time, not information." — Mechanism, falsifiable, disbelieved by CRM incumbents.
Common kernel patterns
Unlock idle supply
Remove trust barrier
Reduce coordination cost
Automate expert task
Collapse distribution cost
Make invisible visible
Uber — unlock idle vehicles via GPS matching
Airbnb — unlock unused homes via trust infrastructure
Stripe — remove developer friction in payments via API
Notion — collapse the coordination cost of knowledge work
Mechanism
Why incumbents dismiss it
Why it scales
Kernel strength
—
Weak
Moderate
Strong
Structural
Enter your kernel insight above.
Kernel test checklist
◈
Open Live Thesis to see your startup logic assembled →
kernel expressed as product
Step 6 — Express the product
Product Expression
The first product that proves the kernel works in practice.
The product expression is the first proof that the kernel works. It does not need to be the final form of the company — it needs to demonstrate the mechanism in the simplest possible way. The product is not the company. The kernel is the company.
Guiding questions
- What is the simplest product that demonstrates the kernel?
- What must the product enable that was impossible before?
- Who is the first customer and what do they experience?
- What moment tells you the kernel is working?
product creates system
Step 7 — Design the flywheel
Flywheel / System
The reinforcing loop that makes the company compound without linear reinvestment.
A flywheel is a reinforcing loop where growth creates the conditions for more growth — without proportional reinvestment. Not every business has a flywheel. Those that do are structurally different from those that don't. If you can't name what gets better as you grow, you may have a business, but not a compounding one.
Guiding questions
- What improves automatically as adoption increases?
- Is there a data loop, network effect, or scale economy here?
- Does the second customer cost less to acquire than the first?
- What would make this flywheel stop spinning?
Flywheel loop preview
Product
→More users
→Data improves
→Product improves
↺system builds moat
Step 8 — Define defensibility
Defensibility
The compounding advantage that competitors cannot replicate from zero.
Defensibility is not a moat you build defensively — it is a structural consequence of the kernel working at scale. The most durable forms of defensibility (data, network effects, switching costs) grow automatically as the company grows. If your advantage doesn't compound with scale, it is a head start, not a moat.
Guiding questions
- Why will this become structurally hard to compete with at scale?
- What advantage grows automatically as the company grows?
- What would a well-funded competitor need to replicate — and why can't they?
timing enables opportunity
Step 9 — Explain the timing
Timing
Why the opportunity exists now. What changed that makes the kernel viable today and not five years ago?
Timing explains why the kernel is viable now and wasn't two years ago. Most failed startups had the right idea at the wrong time. The timing section forces you to identify the structural change — a technology, a cost curve, a behavior shift — that makes this moment different from all previous ones.
Guiding questions
- What technological shift enables this?
- What cost curve has crossed a threshold?
- What behavior change has reached critical mass?
- What infrastructure now exists that didn't before?