Start a Company: From Assumption to Validated Business

How to

1. 1List your customer-facing assumptions explicitly: who has the problem, how acute it is, what they do today.
2. 2Schedule direct conversations with real prospective customers, partners, and suppliers — the founder does this personally, not a hired researcher.
3. 3Run problem-focused interviews: ask about their actual behavior and workarounds before you ever describe your solution.
4. 4Identify early adopters with an acute problem, an existing workaround, and a budget — the people most likely to engage an unfinished product.
5. 5Capture facts, not flattery; write down what people do, not just what they say they'd do.

Watch out for

• —Pitching instead of listening — selling your idea in an interview destroys the data.
• —Talking only to friends and people who already agree with you.
• —Treating polite interest as demand; enthusiasm without behavior is not a fact.
• —Stopping discovery once you 'feel confident' — confidence is not validation.

How to

1. 1Write each belief as a testable statement with a clear pass/fail threshold, not a vague hope.
2. 2Rank assumptions by risk: which one, if false, ends the venture? Test that first.
3. 3Choose the cheapest experiment that can falsify the assumption — an MVP, a landing page, a concierge test, or a sprint prototype.
4. 4For a compressed test, frame the big challenge, get an empowered decider in the room, sketch concrete options, and put a real-seeming prototype in front of target customers within the week.
5. 5Pre-commit to what result will change your mind — define what 'pass' and 'fail' look like before you run it.

Watch out for

• —Running experiments you can't fail — vanity tests that confirm what you already believe.
• —Testing low-risk details while the existential assumption stays untouched.
• —Soliciting 'feedback' instead of observing real reactions; people are polite, behavior is honest.
• —Group brainstorming substituting for structured, individual idea generation in a sprint setting.

How to

1. 1After each experiment, write the single thing you now know that you didn't before — and the behavior that proves it.
2. 2State the decision that learning enables: persevere, iterate, or pivot.
3. 3Track uncertainty falling over time, not just activity rising.
4. 4Use a sprint's customer reactions as concrete data to decide whether to commit, not as a vote on your taste.

Watch out for

• —Confusing data you like with data that's true.
• —Learning something and then not acting on it because of sunk cost.
• —Mistaking a single enthusiastic customer for validated learning across a segment.

How to

1. 1Anchor the offer to the customer's single top priority as discovery revealed it, and quantify the value where you can.
2. 2Write the unique value proposition in the customer's language — benefit first, mechanism second.
3. 3Design the wrapper deliberately: premium, free, or discount packaging changes who responds and at what quality.
4. 4Add a clear call-to-action, handle the obvious objections, and consider recurring/subscription models to get paid more than once.
5. 5Test the offer's pull as a hypothesis — willingness to pay is a fact you gather, not a number you assume.

Watch out for

• —Listing features the team is proud of instead of the benefit the customer cares about.
• —Underpricing out of fear; a price too low can signal low value and starve your economics.
• —Stacking so many offers that the core promise gets muddy.
• —Promising value the product can't yet deliver to the early adopter.

How to

1. 1Focus on a single bounded, dominable beachhead segment and persona rather than chasing many at once.
2. 2Identify your market type — existing, re-segmented, new, or clone — because it governs strategy and how fit will look.
3. 3Watch for the signals: customers using it without prompting, returning, paying, and referring others.
4. 4If the signals aren't there, iterate or pivot the product or model — don't push harder on distribution yet.
5. 5Confirm the problem is urgent and widespread enough that the addressable market can support your goal.

Watch out for

• —Declaring fit on the strength of a few friendly customers.
• —Spreading across multiple segments so you dominate none.
• —Mistaking a great product in a weak market for fit — market quality dominates.
• —Pouring money into acquisition before the pull is real.

How to

1. 1Map who buys and how — the acquisition channels and conversion funnel that reach prospects efficiently.
2. 2Document a repeatable sales process and confirm a second, third, fourth customer come through it without you reinventing the sale.
3. 3Compute unit economics: lifetime value relative to acquisition cost and the payback period; confirm they're viable.
4. 4Set pricing and monetization on value, including recurring revenue where it fits.
5. 5Confirm activation and retention — that the first valuable experience connects to the promise and that customers keep using it.

Watch out for

• —Calling a hand-sold pipeline 'repeatable' because you closed it.
• —Acquisition costs that quietly exceed lifetime value while top-line revenue grows.
• —Confusing one strong channel with a scalable distribution system.
• —Ignoring churn — retention failures invalidate the model no matter how good acquisition looks.

How to

1. 1Track runway explicitly: how many months of search your cash buys at current burn.
2. 2Minimize fixed commitments — overhead, headcount, capital — and substitute sweat for cash where you can.
3. 3Ensure each transaction yields real cash profit, not just revenue.
4. 4Allocate profit first, then operate within the remainder, rather than hoping profit appears.
5. 5Defer scaling spend until model validation is evidenced, not assumed.

Watch out for

• —Hiring ahead of validated demand to feel like a 'real' company.
• —Vanity revenue that loses money on every unit.
• —Confusing cash in the bank with profit — timing can mask insolvency.
• —Premature scaling — the classic killer the development books warn against.

How to

1. 1Verify who is genuinely a founder — by commitment, sacrifice, relationship, and skills fit — versus an advisor or early employee.
2. 2Confirm the team's combined skills cover the venture's critical needs before adding more founders.
3. 3Decide the split method (equal or unequal via a scorecard) and discuss it in percentages, openly, while trust is high.
4. 4Put everyone on a vesting schedule with a cliff; 'set it and forget it' once it's fair.
5. 5Hire for specific strength, not absence of weakness, and build a habit of debating issues openly then committing to the decision.

Watch out for

• —Deferring the equity conversation because it's awkward — it gets harder, not easier.
• —Granting full equity with no vesting and no protection on departure.
• —Bringing on a 'cofounder' who is really a contractor.
• —A team that avoids conflict instead of surfacing and solving issues.

How to

1. 1Confirm you have fit and a validated model before adding scale spend — these precede growth in the sequence.
2. 2If pursuing dominance: design for distribution and network effects, accept some inefficiency for speed, and cross the chasm by dominating a beachhead before broadening.
3. 3If pursuing deliberate smallness: optimize for profitability, low overhead, and owner autonomy rather than headcount.
4. 4Build operational scalability deliberately — systematize core processes so results don't depend on you personally.
5. 5Establish vision alignment and a regular execution pulse so a growing team stays focused on a few priorities.

Watch out for

• —Scaling on the strength of fit alone, before economics are proven.
• —Adding process and structure all at once instead of incrementally as needed.
• —Chasing growth that destroys the margins or the freedom you started for.
• —Letting culture drift as you hire — it must be transmitted deliberately as you scale.

How to

1. 1Write your own definition of success — exit, dominance, durable profitability, or freedom — before you choose a growth and funding path.
2. 2Build the business as a system that can run without you if independence is the goal — work on the business, not only in it.
3. 3Invest in the founder/CEO capacities the corpus emphasizes: transparency, resilience, and the judgment to make lonely decisions.
4. 4Keep the validated-model discipline as the engine — survival follows a proven model, not a hopeful one.
5. 5Sustain cash discipline and team cohesion as the two ongoing enablers of staying alive long enough to win.

Watch out for

• —Adopting someone else's definition of success and optimizing for a future you don't actually want.
• —Building a business that is really a job that depends entirely on you.
• —Treating survival as guaranteed once you have early traction — the hard problems are recipe-less and continuous.
• —Letting the end-goal choice stay implicit until funding and growth decisions have already locked you in.