Startup idea validation: Startup idea validation is the process of gathering evidence that strangers will pay to solve a specific problem, before committing months of building. It is distinct from getting an opinion: an opinion costs the other person nothing, while validation looks for behavior that already has a cost attached, such as people paying for a workaround. A validation process is only meaningful if it can end in "no."
Costly signal: A costly signal is evidence about demand that carried a real cost for the person giving it, which is why it is hard to fake and worth trusting. Someone paying for a clumsy workaround, switching tools, or spending hours solving a problem manually is a costly signal, because they had something to lose. Costly signals predict future demand far better than free encouragement does.
Cheap signal Read more: How to Validate Your Startup Idea in 2026 (Before You Waste 6 Months Building It)
Cheap signal: A cheap signal is approval that cost the person giving it nothing, so it carries almost no information about whether anyone will actually pay. "Cool idea," a friend's thumbs-up, or a chat assistant calling your concept promising are all cheap signals, because saying yes was free and risk-free. Founders get into trouble when they treat cheap signals as if they were costly ones.
Costly signal Read more: How to Validate Your Startup Idea in 2026 (Before You Waste 6 Months Building It)
Kill criteria: Kill criteria are predefined conditions that, when met, mean an idea should be stopped rather than reworked. Setting them before you research protects you from moving the goalposts once you are emotionally attached. A common set: a strong free competitor exists, no differentiation is possible, the market is too small, the idea cannot be monetized, or development cost exceeds the value; when two or more are true, the honest call is to stop.
Read more: How IdeaDose Actually Evaluates Your Idea (And Why You Can Trust the Verdict)
GO / RISKY / KILL verdict: GO, RISKY, and KILL are the three possible outcomes of a threshold-based idea evaluation: GO when no kill criteria are triggered, RISKY when exactly one is, and KILL when two or more are. The point of a three-way verdict is that KILL is a genuine, reachable outcome, not a rounding error. A tool that can only ever land on GO or "promising" is not producing a verdict.
Verdict: A verdict is a committed decision about whether to build, pause, or stop, as opposed to advice that leaves the decision open. Analysis that ends in "here are some options" or "shall we explore further?" is not a verdict, because it hands the hard part back to you. A useful verdict names the outcome and the specific evidence behind it, and it accepts that the outcome could be negative.
False GO (false positive validation): A false GO is a validation result that tells you to build when the real market would have said no, usually because the check was never allowed to fail. It is the most expensive error in early-stage building: a false KILL costs you one idea, but a false GO costs you the months and money you pour into it afterward. False GOs come from cheap signals, confirmation bias, and tools optimized to be encouraging rather than accurate.
Confirmation bias (in idea validation)
Confirmation bias (in idea validation): Confirmation bias in idea validation is the tendency to notice evidence that supports an idea you already want to build and to discount evidence against it. It is dangerous because it feels like research: you genuinely searched, you just searched for reasons to say yes. The countermeasure is to set a disqualifying threshold in advance and to use an outside check that is allowed to reject the idea.
False GO (false positive validation)Read more: Anthropic Warns: Ask AI to Validate Your Startup Idea and It Will Just Agree
Decision paralysis (analysis paralysis): Decision paralysis is the state of being stuck between building, pivoting, and stopping because no source of input will commit to an answer. It is common after a founder has asked several tools and advisors that each reframe the risk without ever saying no. Relief usually comes not from more analysis but from a single grounded read that is willing to reach a definite conclusion.
Problem cost (cost of the problem): Problem cost is how much a problem already costs the people who have it, measured in money, time, or effort they are spending today to cope with it. A problem with high cost pulls people toward any working solution, while a low-cost problem lets them comfortably ignore it for another six months. Checking problem cost is more predictive than asking whether people like your idea.
Minimum viable product (MVP): A minimum viable product (MVP) is the smallest version of a product that can deliver its core value to real users. An MVP is a way to test execution and usability, not a substitute for validating demand: building one still costs weeks or months, so a silent MVP often means the idea itself, not the build, was the untested assumption. Demand is cheaper to test before an MVP than after it.
Sunk cost fallacy (in building)
Validation vs distribution: Validation asks whether an idea is worth building; distribution asks whether you can reliably reach the people who would buy it. They are different failure modes: a validated idea with no distribution stalls, and great distribution behind an unvalidated idea just spends money faster. Treating a distribution problem as a validation problem, or the reverse, sends founders fixing the wrong thing.
Demand signal: A demand signal is observable evidence that real people already want a problem solved, such as recurring complaints, "is there a tool that does X" questions, or people paying for imperfect alternatives. The strongest demand signals are specific and recent rather than general and old. Volume of demand signals matters: a handful of vague mentions is not the same as dozens of frustrated, dated posts.
Willingness to pay: Willingness to pay is whether people will actually spend money to solve a problem, as opposed to merely liking the idea of a solution. It is the signal most founders skip, because interest in a free product tells you almost nothing about a paid one. Pre-orders, existing spend on workarounds, and honest "no" answers to a direct price question all measure willingness to pay better than waitlist enthusiasm does.
Duct-tape solution (workaround): A duct-tape solution is the clumsy workaround people already use to cope with a problem, such as a janky spreadsheet, a manual routine, or a paid virtual assistant. It is one of the strongest validation signals because it is a costly signal: the person is already spending money or effort, which means they are pre-sold on a better answer. Finding duct-tape solutions is more reliable than finding people who say they would use your product.
Read more: How to Validate Your Startup Idea in 2026 (Before You Waste 6 Months Building It)
Market saturation (red ocean): Market saturation, sometimes called a red ocean, is a market already crowded with established players competing on the same value, often including strong free options. Saturation is not automatically fatal, but it raises the bar: entering one usually requires a specific, defensible wedge rather than a general "better version." A common validation error is calling a saturated market "attractive" while naming the very competitors that make it hard.
Sunk cost fallacy (in building): The sunk cost fallacy in building is continuing with an idea because of the time and money already spent, rather than because the evidence still supports it. It is why a founder who has shipped an MVP finds it hardest to accept a KILL signal: the effort already invested feels like a reason to keep going. The evidence-based question is not "how much have I put in," but "would I start this today knowing what I now know."
Minimum viable product (MVP)
Accountability check (accountability cofounder): An accountability check is an outside read that is allowed to tell you the idea itself is the problem, not just the execution, and that holds you to a decision you might otherwise avoid. It differs from a brainstorming partner, which is designed to agree and keep the conversation going. The value is structural: a tool built to be agreeable cannot also be the thing that tells you to stop.
Grounded evaluation: A grounded evaluation is one whose conclusions trace back to real, checkable sources, such as live competitor listings, current pricing, and community posts, rather than to plausible text generated from training data. Grounding matters because a confident answer can be produced without verifying a single competitor or demand signal. A grounded check shows what it found and marks weak areas as low-confidence instead of guessing.
Allowed to fail: "Allowed to fail" describes a check designed so that a negative answer is a real, reachable outcome, not something the tool structurally avoids. A validation tool that is optimized to be helpful and agreeable is, by design, not allowed to fail your idea, so its approval carries little weight. The test for any idea-checking tool is simple: has it ever, on real data, told someone no.
Idea-market fit: Idea-market fit is whether the core idea addresses a real, costly problem for a reachable group of people, established before you build. It comes earlier than product-market fit, which is measured only once a working product is in users' hands. Confirming idea-market fit first is what keeps a founder from discovering, after an MVP, that the problem was never worth paying to solve.

Ideadose

Startup Idea Validation Glossary