Before you start
Free · self-guided · nothing is sent anywhere
The Builder's Blueprint
You don't need a new idea. You need to find the one already hiding in what you know, what you've done, what you're good at, what you've suffered — and what you've won.
This takes about 35 minutes and it is real work, not a personality test. There is no score at the end — only your own map.
Your answers save in this browser as you type. Close the tab and come back — nothing is transmitted, and you can verify that with your network tab open.
This browser won't let the page save. You can still finish in one sitting, but don't close the tab.
K · Knowledge — bird-in-hand
What do you know deeply?
From school, from a job, or from your own obsessive curiosity. Don't filter for "marketable" yet — just true.
E · Experience
What have you actually done?
Jobs, projects, seasons that shaped you — including the ones that felt like detours at the time. Not a résumé.
S · Skills
Who has asked you for help — and with what?
Not "what are you good at." Name a real person from the last year or two, and what exactly they came to you for. What looks easy to you and hard to everyone else?
P · Pains — the letter the market drops
What frustrates you — and what have you already fixed for someone else?
Both halves matter. The second is usually the richest and most overlooked: a real problem you already solved for a specific person.
Your own pains
Pains you've already solved for other people
V · Victories — the letter the market also drops
What have you actually pulled off?
A result you produced, a problem you beat, a number you moved. Don't be modest; be specific.
Wins you've booked
Wins others would pay to repeat
The Five Whys · Toyota Production System
Why does each pain actually happen?
The first answer is almost always a symptom. Keep going — the opportunity usually sits one or two layers under it. A real problem can have more than one cause, so add a branch wherever two things are both true.
Taiichi Ohno developed this at Toyota. It is a good way to stop at the first plausible answer less often — not a guarantee you have found "the" root cause. Critics rightly point out that forcing a single path is its main weakness, which is why this version lets you branch.
Go three "whys" deep on at least one pain before moving on. That third layer is where this stops being obvious.
The crossing matrix
Where does your raw material meet a real pain?
Mark how strongly each root-caused pain overlaps what you know, what you've done, and what you're good at. The strongest crossings are where you are credible and the problem is real at the same time.
Reality check · the external pass
Would anyone outside your own head pay for this?
People systematically overvalue what they build themselves — psychologists call it the IKEA effect, and it is well replicated. That is exactly why this step exists: your own enthusiasm is not evidence.
For each strong crossing below: has anyone, anywhere, ever paid to solve a version of this — even badly? Who do you already know who has this exact pain? And who do you know who could introduce you to more of them, or put you in front of a room full of them?
Your options
Name one to three — never force a third.
One block per surviving crossing. Specific beats broad here: research on goal setting is unusually consistent that vague targets underperform, and "everyone who struggles with X" is not a target.
Three is the maximum — focus is the point. If a fourth idea is strong, it belongs in your next round, not this one.
Focus
Which one first?
You choose — focus is the whole game. This becomes your ONE goal. Everything after this step is about that one thing only.
The riskiest assumption
What has to be true — that you're least sure of?
Every plan rests on something unproven. Name the one that would sink this if it were false, then design the cheapest thing that could prove you wrong this week. Aim to learn, not to be right.
Of everything on this page, this step has the strongest evidence behind it: teaching founders to state falsifiable hypotheses and test them improved decisions in randomised trials of several hundred firms.
Implementation intentions
Three actions — when, and where.
Not "keep thinking about it." Use the format "When [trigger], I will [specific action]" — naming when and where measurably raises the odds you actually do it. Make at least one of these a person you named in the reality check.
Your map
You just found the one thing.
This is yours. It never left your browser, and you can take it whether or not you ever talk to me.
Which is closer to true?
"I want to build this with someone."
Apply with this map →Applications open soon — the founding cohort is being placed by hand.
The evidence behind this
What's actually known — and what isn't.
Each move in this worksheet draws on published research. That is a claim about the components, not about the outcome. Here is the honest version.
STARTING FROM WHAT YOU ALREADY HAVE
Research on expert entrepreneurs finds they often reason from means they already hold — who they are, what they know, whom they know — rather than from a fixed goal (Sarasvathy, Academy of Management Review, 2001). A meta-analysis of roughly 9,900 ventures links that means-driven orientation to venture performance (Read, Song & Smit, Journal of Business Venturing, 2009). Both are correlational. Neither shows that reasoning this way causes success.
ASKING WHY REPEATEDLY
The five whys was developed by Taiichi Ohno at Toyota. It is genuinely useful for getting past the first plausible answer — but its popularity is not the result of evidence that it works, and a review in BMJ Quality & Safety criticises it for forcing a single line of reasoning toward a single root cause (Card, 2017). That critique is why this version lets you branch. Separately, asking "why is this true?" is rated a moderate-utility learning technique in a review of ten study strategies (Dunlosky et al., 2013) — moderate, not high.
ASKING FOR SPECIFICS INSTEAD OF SELF-RATINGS
Across 22 meta-analyses, how people rate their own ability correlates with their actual performance at about r = .29 — and narrow, concrete self-assessments hold up far better than broad ones (Zell & Krizan, Perspectives on Psychological Science, 2014). That is why every question here asks you to recall something specific and checkable, and why there is no score.
BEING SPECIFIC ABOUT WHO
Specific, challenging goals outperform vague ones across decades of studies (Locke & Latham, American Psychologist, 2002). The effect shrinks as tasks get more complex, and finding a business is about as complex as tasks get — so treat this as a reason to be precise, not as a promise.
TESTING THE RISKIEST ASSUMPTION
This is the strongest evidence on the page. In randomised controlled trials of Italian startups, founders taught to frame explicit theories, state falsifiable hypotheses and test them made better decisions than a control group using ordinary search heuristics (Camuffo, Cordova, Gambardella & Spina, Management Science, 2020), replicated across 759 firms (Strategic Management Journal, 2024).
PLANNING WHEN AND WHERE
Framing a next action as a specific if-then plan improves follow-through (Gollwitzer & Sheeran, 2006). A 2024 meta-analysis of 642 tests by the same authors found substantial publication bias in that literature and smaller real effects, so treat it as a good habit with real support behind it rather than a large guaranteed gain.
WHY THE LAST STEP IS AN OUTSIDE CHECK
People systematically overvalue things they made themselves — the IKEA effect (Norton, Mochon & Ariely, 2012; replicated at meta-analytic scale). Your own map is not exempt. That bias is the reason this asks whether anyone has ever actually paid, and who you already know.
WHAT THIS IS NOT
This is not a validated instrument. There is no score, no norms, no psychometrics — nothing here has been tested as a whole. And you should know that no ideation worksheet has been shown to make anyone succeed, including the famous ones: the Business Model Canvas and Lean Startup are far more studied than this and have close to zero controlled evidence behind them. One study followed 623 people starting businesses for six years and found formal planning didn't move performance at all.
What you're getting is a set of good moves, in a sensible order, that forces you to be specific. That's worth something. It isn't a guarantee, and anyone selling you one is lying.