Every other domain in this course — positioning, customer discovery, business models, growth — sits on top of a handful of ideas that most founders think they already understand and mostly don't. Get these wrong and everything downstream is built on sand: you'll validate the wrong problem, chase a business plan when you actually have a startup, or mistake enthusiasm for evidence. Use this guide as the reference you come back to whenever a mission mentions a term you can define but can't quite defend.
Problem
A problem is a struggle a customer has with a job they're trying to get done — and it's only worth building for if it's painful enough that the customer is already doing something about it, even a bad something.
This is the foundation Rob Fitzpatrick builds The Mom Test on: you can't find real problems by asking people what they think or what they'd want. You have to ask about specifics in the past. "Would you use an app that organizes recipes?" gets you a hypothetical, feel-good "yes." "What's the last cookbook you actually bought, and why?" gets you a fact. Fitzpatrick's litmus test for a real problem is behavioral evidence of struggle: workarounds, spreadsheets duct-taped together, money already being spent on a worse solution. Fitzpatrick also draws the line between painkillers and vitamins: a "must-solve-right-now" problem versus a "nice-to-have" — a problem people find mildly interesting isn't the same as a problem they're organizing their day around.
Micro-example: a freelance bookkeeper who exports bank data to Excel by hand every Friday night, and has done it for three years, has a problem. A freelance bookkeeper who says "it'd be nice if this were automated" when you ask, but has never looked for a tool, has an opinion.
Watch out: treating a complaint as a problem. People complain about lots of things they'll never pay to fix — weather, traffic, their own procrastination. The signal isn't the complaint, it's the existing, clumsy attempt to solve it.
→ practice this in the painful-problems mission.
Product
A product is a repeatable solution to a problem a specific group urgently wants solved — one that works without you, the maker, standing next to the customer explaining how to use it.
This is a higher bar than "thing I built." A demo you drive by hand for one customer is a prototype, not a product; a feature that only works because you personally onboard every user isn't repeatable yet. Peter Thiel's framing in Zero to One is useful here at the extreme end: the best products aren't incremental tweaks on what exists (1 to n), they're a genuinely new answer to a problem (0 to 1) — going from "no good way to do X" to "a way to do X." Most products you'll build won't be that dramatic, but the test still applies at small scale: does this solve the problem repeatably, for people who aren't you and weren't coached by you?
Micro-example: you personally reconciling a client's books in a shared spreadsheet is a service. A tool that ingests their bank CSV and reconciles it automatically for anyone who signs up is a product.
Watch out: confusing "I built something" with "I built a product." A pile of features nobody asked for, however well-engineered, isn't a product until it's aimed at a validated problem.
→ practice this in the what-is-a-product mission.
Startup vs. Business
A startup is a temporary organization searching for a repeatable, scalable business model under extreme uncertainty. A business is an organization that already has one: it sells something for reliably more than it costs to make and deliver, and its main job is executing that known model well.
Eric Ries is precise about this distinction in The Lean Startup: "a startup is a human institution designed to create a new product or service under conditions of extreme uncertainty." The key phrase is what it excludes — a business that's an exact clone of an existing one, with a known model, known pricing, and a known customer, isn't a startup no matter how small or new it is, because a loan officer could underwrite its prospects with confidence. A startup, by contrast, is still searching. Rob Walling's Start Small, Stay Small draws a related but different line: between a bootstrapper, who wants a company of real size built through reinvested profit, and a micropreneur, who wants a lifestyle business run by one or two people that stays intentionally small. Both can be run without venture money; neither is automatically a "startup" in Ries's sense — a lifestyle SaaS product in a validated niche, competently run, can be a business from day one.
Micro-example: a freelancer who opens a second identical bookkeeping practice in a new city is running a business — the model, pricing, and customer are already known. A freelancer building a piece of software that automates bookkeeping for a customer segment nobody has proven will pay for it yet is running a startup, even if it's a company of one.
Watch out: calling every early-stage company a "startup" out of habit. If you already know your customer, your price, and that people will pay it, you're operating a business and should be managed like one — process, predictability, and execution matter more than experimentation.
→ practice this in the startup-vs-business mission.
Founder-Market Fit
Founder-market fit is the advantage a founder has in a specific market — from lived experience, distribution, or hard-won skill — that competitors can't easily copy.
Scott Kupor's Secrets of Sand Hill Road names this directly as something investors screen for: once a VC believes the market opportunity is real, the next question is "why you?" — why should they back this founder instead of waiting for the next team to walk in with the same idea. Kupor's standout example is Martin Casado, whose PhD thesis effectively was the technology behind the company he later founded (Nicira, acquired for $1.25B) — about as strong a founder-market fit story as exists. But the advantage doesn't have to be a PhD: it can be a founder who lived the exact pain point for years, has a distribution channel competitors don't (an audience, a network, a domain reputation), or has a rare combination of skills — like Jack Dorsey's ability to get Square onto The Oprah Winfrey Show and into a J.P. Morgan partnership through sheer personal credibility. Rob Walling's niche-picking advice in Start Small, Stay Small is the bootstrapped version of the same idea: the niches worth choosing are usually ones you already understand from the inside.
Micro-example: two founders both want to build scheduling software for countertop contractors. One spent five years running a countertop installation business; the other found the niche on a market-sizing spreadsheet. All else equal, the first has founder-market fit — they already know the workflow, the vocabulary, and probably a few dozen future customers by name.
Watch out: treating founder-market fit as a fixed trait you either have or don't. It's a story you build evidence for — through customer conversations, a track record, or distribution you've earned — not a credential you can claim by assertion.
→ practice this in the who-is-it-for mission.
Vision
Vision is the future state you're betting on — stable enough to actually steer by, but held loosely enough to survive contact with evidence.
Ries's model in The Lean Startup separates three layers that people constantly conflate: vision, strategy, and product. The vision is the destination and rarely changes. The strategy — your business model, your product roadmap, your beliefs about competitors and partners — gets you there, and can change; a strategy-level change is what Ries calls a pivot. The product is the day-to-day output of that strategy, and changes constantly through small optimization ("tuning the engine"). Confusing these layers is a common failure mode: founders either treat their vision as changeable (and end up with no throughline at all — Thiel's "indefinite optimism," expecting the future to be better without a specific plan to make it so) or treat their strategy as sacred (and refuse to pivot even when the evidence says the current approach won't reach the vision).
Micro-example: "make small business accounting effortless" is a vision. "Do it by automating bank reconciliation for freelance bookkeepers first" is a strategy — and if that channel proves too small, switching to target dental practices instead is a pivot, not a betrayal of the vision.
Watch out: mistaking a pivot for failure, or mistaking stubbornness for vision. Ries is explicit that most successful startups pivot at least once — the vision survives; the path to it doesn't have to.
Risk
Risk, in this context, means the specific assumptions — desirability, viability, feasibility — that would kill the venture if they turned out to be wrong, ranked so you test the deadliest ones first.
Testing Business Ideas (Bland & Osterwalder) organizes nearly the entire book around this three-way split. Desirability risk asks whether anyone actually wants this — will customers care enough to switch. Viability risk asks whether the economics work — can you charge enough, cheaply enough, to build a real business. Feasibility risk asks whether you can actually build and deliver it — technically, operationally, legally. Founders' instinct is almost always to attack feasibility risk first, because it's the most comfortable one (it's the thing you already know how to do — write code, build a supply chain). But feasibility risk is rarely what kills startups; desirability risk usually is. The discipline is to identify your riskiest, least-evidenced assumption in any of the three categories and test that one before you build anything expensive.
Micro-example: before writing a line of code for a bank-reconciliation tool, the riskiest assumption isn't "can we build the CSV parser" (feasibility, well understood) — it's "will freelance bookkeepers actually switch from their existing spreadsheet habit" (desirability, unproven). That's the one to test first, with a landing page or five customer interviews, not a sprint of engineering.
Watch out: working down the risk list in order of comfort instead of order of danger. The point of ranking risk is to spend your scarcest resource — time before you run out of money — on the assumption most likely to end the venture.
Evidence and Iteration
Evidence is observed customer behavior — payments, sign-ups, repeated use — as distinct from opinions, compliments, or your own enthusiasm. Iteration is the loop of shipping something small, learning from how real people actually use it, and revising — the mechanism that turns evidence into a better product instead of a better story.
The clearest description of why this matters comes from Alberto Savoia's The Right It: most ideas die not because the market rejects them, but because they were never actually market-tested — they were incubated entirely in what Savoia calls Thoughtland, an imaginary place where an idea gets refined purely through opinions (yours, your team's, a focus group's) with nobody having any skin in the game. Opinions, Savoia argues, aren't data; they're guesses with no cost attached to being wrong. Testing Business Ideas formalizes the same instinct into a strength hierarchy: opinions are weaker evidence than facts, what people say is weaker than what people do, and a lab setting is weaker than real-world behavior. A survey response ("I'd definitely use that") sits near the bottom; a stranger paying with their own card sits near the top. Ries's Build-Measure-Learn loop, from The Lean Startup, is iteration made procedural: build the smallest thing that lets you measure a real customer reaction, measure it, learn whether to persevere or pivot, and repeat — minimizing the total time through the loop is the actual skill, not any single pass through it.
Micro-example: "12 people in a Slack poll said they'd pay for this" is weak evidence (opinion, hypothetical, no cost). "5 people pre-paid $20 for early access after seeing a two-minute demo" is strong evidence (fact, action, real stakes) — even though the sample is smaller.
Watch out: mistaking enthusiasm for evidence, especially from friends, family, or people who like you and want to be encouraging. The Mom Test exists specifically because your mother will tell you your idea is great regardless of the idea.
→ practice this in the mindset-of-evidence mission.
Product-Market Fit
Product-market fit is the state where a specific market pulls the product out of your hands — retention holds, word of mouth compounds, and growth stops feeling like pushing a boulder uphill.
The term, as Ries recounts in The Lean Startup, was coined by Marc Andreessen: "in a great market — a market with lots of real potential customers — the market pulls product out of the startup." Ries's contribution is making it measurable instead of vibes-based: because fit is always fit with a specific customer segment, and because growth engines (viral, sticky, paid) have their own quantifiable metrics, you can track whether you're getting closer to fit by watching the trendline, not the raw number. His two-company example makes the point sharply: a startup compounding at 5% growth but climbing steadily month over month (0.1% → 0.5% → 2% → 5%) is in a fundamentally better position than one flat-lined at a higher but static 10% — because the first is tuning its way toward fit and the second has already found its ceiling. Retention is the clearest single signal: if people who try the product keep using it without you chasing them, that's fit; if you have to keep re-selling existing customers, you don't have it yet.
Micro-example: a freelancer tool where 60% of signups are still active and paying at month three, and a third of new signups arrive via referral with no ad spend, is showing real product-market fit signals. The same tool with 60% signups but an 8% month-three retention rate is not — no matter how good the initial launch numbers looked.
Watch out: declaring product-market fit off a good launch week. Ries's own line is blunt: "if you are asking, you're not there yet." Fit shows up in trends over months, specifically in retention, not in a single spike of sign-ups.
Go deeper
- Rob Fitzpatrick, The Mom Test — for running customer conversations that surface real problems instead of polite lies.
- Peter Thiel, Zero to One — for the 0-to-1 framing of what makes a product genuinely new, and definite vs. indefinite optimism as it applies to vision.
- Eric Ries, The Lean Startup — for Build-Measure-Learn, the startup/business distinction, vision/strategy/product, pivot vs. persevere, and product-market fit.
- Alberto Savoia, The Right It — for the Law of Market Failure, Thoughtland, and why opinions aren't evidence.
- Rob Walling, Start Small, Stay Small — for bootstrapper vs. micropreneur, and why niches are the way in.
- Scott Kupor, Secrets of Sand Hill Road — for founder-market fit from an investor's point of view.
- David Bland & Alexander Osterwalder, Testing Business Ideas — for desirability/viability/feasibility risk and the evidence-strength hierarchy.