创业验证 — 完整指南
我们发布的关于在构建前验证创业想法的所有内容的中心枢纽。为首次创业者、独立开发者和任何盯着空白 Notion 文档的人而写。
最后更新 · 2026年8月31日
创业验证,一句话总结
- 创业验证是一种有纪律的过程——在投入数月构建之前,测试你的想法背后的假设。
- 验证回答四个问题:问题是真实存在的吗?你能触达客户吗?他们会付费吗?你能交付吗?
- 输出的不是确定性,而是证据——每条声明都标注来源,每个分数都能追溯到一个具体规则。
- 验证比构建便宜:一次问题访谈、一个落地页测试和一次付费试点,2-6 周就能跑完。
- 想要的输出不是分数,而是找到那个如果错了就让整个计划失效的、最关键的那个假设。
验证如何串联起创业的其他环节
验证不是单一话题,而是把创业其他环节串起来的脊梁。下面的链接展示了创始人通常按怎样的顺序走过这条链。
什么是创业验证？
创业验证是一种有纪律的过程 —— 在投入数月构建产品之前，测试想法背后的假设：问题、客户、付费意愿，以及你执行的能力。
它不是单一事件。它是一连串便宜、快速的实验，把模糊的直觉变成可行动的证据。目的不是证明想法好，而是尽早、尽量便宜地发现它可能在哪些地方失败。
跳过验证的代价
大多数失败的创业公司并非在发布时失败。它们在键盘前失败 —— 早了好几个月，那时创始人选择构建一个没人要求的产品。你用于编码的四个月，正是你无法用来学习应该构建什么的那四个月。
时间
几个月的构建付诸东流
大多数创始人花 3-6 个月构建产品，才发现解决了一个没人要求解决的问题。验证可以在几天内发现这一点。
金钱
资金花在未经测试的赌注上
每构建一个月，就烧掉一个月的跑道，押在一个可能没有客户的 idea 上。验证会缩小赌注的规模。
信心
由证据支撑的决策
经过验证的想法仍然可能失败。但失败是市场的失败，不是猜测的失败。你会从中学到东西，而不是归咎于执行。
哪个工具对应你的问题？
这四个工具都跑同一个 Startup MRI 引擎。区别在于每个工具首先帮你回答哪个问题。选最贴合你当前处境的那一个，可以随时切换。
如果你的问题是……
……我应该先验证哪个假设？
指出本周应该爬的台阶。7 级验证阶梯把 Startup MRI 的每个维度映射到它最便宜的行为信号测试，按创始人明天就能跑的顺序排列。
查看 →如果你的问题是……
……这个点子配得上我的执行能力吗？
把点子放进 2×2 的「想法强度 × 执行能力」网格。每个象限对应一个明确的下一步——构建、找联合创始人、重新定位或放弃。
查看 →如果你的问题是……
……这个点子配得上多少分？
返回七个维度上的 0–100 评分，并给出四个分数段和四种分数模式，让你能够解释这个分数，而不是只看到一个数字。
查看 →如果你的问题是……
……本周我应该做什么？
返回完整的 Startup MRI 报告：唯一的核心假设、由最高风险维度派生的 7 天验证行动清单，以及一份 4 周首批客户计划。
查看 →
循序渐进的阅读路径
从顶部开始，或直接跳到最接近你想法的产品类型。每一步都链接到一篇已发布的指南，并补充一个不同的验证视角。
初学者路径
先理解问题与客户，再选择产品类型。完成前五步后，再进入与你想法匹配的实验。
从这里开始 — 心态
Why Founders Build Before They Validate
为什么这么多创始人在拿起电话之前就拿起了键盘？先读这篇，理解让我们跳过验证的心理模式。
15 分钟阅读 ·阅读这一步 →
框架
How to Validate a Startup Idea Before You Build
在写一行代码之前，测试问题、客户、商业模式和你执行能力的实用四阶段框架。
17 分钟阅读 ·阅读这一步 →
问题-方案匹配
Problem-Solution Fit: How to Know Your Solution Solves a Real Problem Before You Build It
产品-市场匹配之前的一级台阶：拟定的方案能否为定义明确的客户解决真实、反复出现的问题？五级证据阶梯、六种失败模式,以及围绕替代方案分析而非礼貌性同意构建的 14 天验证框架。
22 分钟阅读 ·阅读这一步 →
清单
Startup Validation Checklist: Before You Build
四个阶段共三十项具体检查。读一遍，然后带着你的具体想法回来。
17 分钟阅读 ·阅读这一步 →
寻找第一批客户
How to Find Your First Customers Before You Build
未来的客户实际聚集在哪里，如何开始一次不是推销的对话，以及一个本周就可以开始的每周计划。
16 分钟阅读 ·阅读这一步 →
付费意愿
How to Test Willingness to Pay Before You Build: Six Experiments, Four Cases, and a Seven-Day Plan
兴趣不是付费。最危险的创业假设是「我会买这个」意味着什么。学习区分礼貌同意与真实金钱的实验。
14 分钟阅读 ·阅读这一步 →
付费意愿验证
Willingness to Pay Validation: The Framework, the Signal Ladder, and Why Interest Is Not Payment
先给出定义,再教你测试的入门指南 —— 从礼貌性话语到真实付款的五级信号阶梯,以及本周就能跑的最便宜实验。
14 分钟阅读 ·阅读这一步 →
产品-市场匹配
Product-Market Fit Validation: How to Know You Have It Before You Announce It
PMF 是被观察到的,不是被宣告的。四个信号告诉你拥有它,七个失败模式告诉你没有,以及一份区分真实留存与付费推广的 30 天验证框架。
22 分钟阅读 ·阅读这一步 →
MVP 验证
MVP Validation: How to Test a Minimum Viable Product Before You Build It
在承诺完整构建之前,如何测试产品的最小可用版本。从最便宜到最强信号排列的六种方法,一个五步框架,以及在创始人注意到之前就出现的七种失败模式。
20 分钟阅读 ·阅读这一步 →
SaaS 专项
How to Validate a SaaS Idea Before You Build It
面向早期 SaaS 与 micro-SaaS 创始人的聚焦版清单——订阅式收入假设栈,以及测试每一项最便宜的实验。
16 分钟阅读 ·阅读这一步 →
AI 专项
How to Validate an AI Startup Idea
在把 AI 能力当作业务之前，测试工作流价值、输出质量、模型层风险和防御性。
21 分钟阅读 ·阅读这一步 →
AI 与 SaaS
AI Startup vs SaaS Startup: How Validation Is Different
比较依赖模型层的软件与传统产品基础设施在验证问题上的不同。
18 分钟阅读 ·阅读这一步 →
Marketplace 专项
How to Validate a Marketplace Startup Before You Build It
测试市场双方、最初的细分切口，以及让有效交易发生的条件。
17 分钟阅读 ·阅读这一步 →
移动应用专项
How to Validate a Mobile App Idea Before You Build It
在构建应用前，验证用户习惯、分发路径、留存信号和应用商店限制。
17 分钟阅读 ·阅读这一步 →
Chrome 扩展专项
How to Validate a Chrome Extension Idea Before You Build It
在发布扩展前，测试浏览器工作流、权限边界、分发渠道和重复使用场景。
16 分钟阅读 ·阅读这一步 →
B2B 专项
How to Validate a B2B Startup Idea Before You Build It
测试问题、采购委员会、设计合作伙伴路径，以及企业客户采用产品所需的证据。
17 分钟阅读 ·阅读这一步 →
API 专项
How to Validate an API Startup Idea Before You Build It
在构建 API 前，测试开发者工作流、集成摩擦、可靠性预期和首个分发路径。
18 分钟阅读 ·阅读这一步 →
获客渠道专项
Distribution Channels Ranked for Solo Founders
十一个面向独立创业者的获客渠道排名——以及一份测试其中两个、不会把人累垮的 4 周前 100 名客户计划。
14 分钟阅读 ·阅读这一步 →
Mom Test
The Mom Test Explained for Solo Founders
Rob Fitzpatrick 的三条规则在独立创业者身上的应用:聊客户的生活、不聊你的产品;问过去的具体细节、别问未来的假设;多听少说。附一份本周就能跑的 5 次访谈脚本。
14 分钟阅读 ·阅读这一步 →
客户访谈问题
Customer Interview Questions for Startup Validation: How to Ask Ones That Produce Evidence
逐题的执行指南:如何向客户询问过去的行为、当前的替代方案、真实的支出和决策语境,以及如何把每个回答转化为更短的待办清单。
18 分钟阅读 ·阅读这一步 →
写代码前的需求验证
How to Validate Demand Before Coding (Without Building the Wrong Thing)
面向非技术背景和独立创始人的写代码前门控:五项证据检查证明可以开始写代码,三个标注示例,以及一条七问规则决定「先写 vs 再验证」。
14 分钟阅读 ·阅读这一步 →
进阶路径
选择与你想法匹配的产品或分发路径。每篇指南测试一种不同风险；当证据改变下一步决策时就停下来。
按主题分类的所有支柱文章
支柱文章按四个编辑分类组织。每个分类回答验证过程中的不同问题。
客户发现
谁有问题？你怎么找到他们？
Customer Research
How to Know When Customer Interviews Are Enough (A Stop-Rule for Solo Founders)
A practical stop-rule for early-stage customer interviews: when to stop, how to recognise the saturation point, why twelve is usually the right number for a single segment, and the four failure modes that keep founders interviewing long after the evidence has stopped arriving — including the difference between code saturation and meaning saturation and the named Guest, Bunce & Johnson (2006) study that produced the empirical baseline.
13 min read
Customer Research
Customer Interview Questions for Startup Validation: How to Ask Ones That Produce Evidence
Customer interview questions for startup validation, organized by what you need to learn — past behavior, current workarounds, decision context, and spending — plus the questions founders should stop asking, an interview-to-decision framework, and a realistic B2B SaaS scenario.
18 min read
Customer Research
The Mom Test Explained for Solo Founders
Rob Fitzpatrick's book, applied to indie founders — the three rules, twelve questions, and a 5-interview script you can run this week to test whether anyone has the problem you think they have.
14 min read
Customer Research
How to Find Your First Customers Before You Build
Where early customers actually gather, how to recognize real Early Adopters vs polite audiences, and a one-week customer-discovery plan you can start this week.
16 min read
验证
在构建之前测试想法
Validation Guide
How to Size a Market You Can Actually Reach: TAM, SAM, SOM, and the Two Numbers That Are Usually Useless
A founder's guide to estimating the market they can actually reach — what TAM, SAM, and SOM actually mean, why two of the three are usually useless for solo founders, how to do a defensible bottom-up sizing in a single afternoon, where to source the data, and the failure modes that produce a market-size claim a partner or investor will not trust.
16 min read
Validation Guide
How to Write a Smoke-Test Landing Page in 60 Minutes
The shortest path from a hypothesis to a real conversion number — page structure, headline formula, the four signal lines, the seven things to leave off, the 60-minute build checklist, and how to read the conversion data at small N.
14 min read
Validation Guide
Fake Door Test: How to Validate Demand Without Writing a Single Line of Code
A complete playbook for the fake door (painted door) test — the cheapest credible demand test in the startup validation toolkit. Includes the original methodology from Alistair Croll & Benjamin Yoskovitz's Lean Analytics, the six-step process, the Unbounce 2024 benchmark that defines what counts as a 'good' smoke-test conversion rate, the two documented case studies (Buffer, Dropbox), the seven failure modes that produce false positives, and what the test cannot prove.
16 min read
Validation Guide
How to Validate Demand Before Coding (Without Building the Wrong Thing)
A step-by-step playbook for non-technical and solo founders who need to know whether anyone will pay for an idea before writing a line of code — including the five pre-code evidence checks, three labeled hypothetical examples, the seven-question rule that decides 'build now vs. validate more,' and the failure modes that lead founders to mistake enthusiasm for demand.
14 min read
Monetization Validation
How to Price a New SaaS Product: Three Decisions, Four Questions, and the Only Pricing Test That Survives Launch
The three pricing decisions a new SaaS founder actually owns (model, value metric, price points), the four Van Westendorp questions that produce real willingness-to-pay data, and the eight-week playbook that turns the answers into a defensible price.
18 min read
Validation Guide
Startup Failure Analysis: How to Read a Failed Startup the Right Way
Startup failure analysis is the discipline of converting documented startup failures into testable assumptions. Five real cases — Quibi, Webvan, Juicero, Homejoy, WeWork — analyzed with a seven-rung framework, then turned into the validation experiments a solo founder can run this week.
26 min read
Validation Guide
Lean Startup Validation: How To Choose, Design and Interpret Validation Experiments
How to use Eric Ries's Build-Measure-Learn loop to pick the next validation experiment — including the five-question decision framework, decision-threshold thinking, three labeled hypothetical examples, and the failure modes that distort the loop before the founder notices them.
17 min read
Validation Guide
Problem-Solution Fit: How to Know Your Solution Solves a Real Problem Before You Build It
Problem-solution fit is the state where a defined customer has a defined problem and a defined solution would meaningfully improve their situation. The five rungs of evidence that prove the fit is real, the six failure modes that show it isn't, and the 14-day validation framework that separates real pain from polite enthusiasm.
22 min read
Validation Guide
MVP Validation: How to Test a Minimum Viable Product Before You Build It
MVP validation is the discipline of testing the smallest useful version of your idea before committing months to a full build. Six methods, a five-step framework, and the seven failure modes that show up before the founder notices them.
20 min read
Validation Guide
Product-Market Fit Validation: How to Know You Have It Before You Announce It
Product-market fit is observed, not declared. The four signals that show you have it, the seven failure modes that show you do not, and a 30-day validation framework that separates real retention from paid acquisition.
22 min read
Monetization Validation
Willingness to Pay Validation: The Framework, the Signal Ladder, and Why Interest Is Not Payment
What willingness to pay actually means, the five-rung signal ladder from polite words to real money, and the Problem → Customer → Value → Price → Payment framework every pricing experiment is a sub-test of — before you write code.
14 min read
Validation Guide
How to Validate an API Startup Idea Before You Build It
The API- and developer-tool-specific tests for technical buyers, integration cost, trust, documentation prototypes, and design-partner pilots — before you write the first endpoint. A practical handbook for API founders, SDK builders, infrastructure product teams, and developer-tool indie hackers.
18 min read
Validation Guide
How to Validate a B2B Startup Idea Before You Build It
The B2B-specific tests for buying committees, procurement, ROI proof, founder-led sales, and the manual pilot — before you write code. A practical handbook for SaaS founders selling to businesses, enterprise software teams, and technical founders.
17 min read
Validation Guide
How to Validate a Chrome Extension Idea Before You Build It
The browser-extension-specific tests for Manifest V3 fit, Chrome Web Store policy, distribution outside store search, willingness to pay, and unlisted pre-launch testing — before you ship a packaged extension. A practical handbook for indie hackers, SaaS founders, AI tool builders, and browser extension developers.
16 min read
Validation Guide
How to Validate a Mobile App Idea Before You Build It
The mobile-app-specific tests for problem, retention, onboarding, distribution, and willingness to pay — before you ship a binary to the App Store. A practical handbook for consumer, productivity, lifestyle, health, education, and local-service apps.
17 min read
Validation Guide
How to Validate a Marketplace Startup Before You Build It
The marketplace-specific tests for supply, demand, liquidity, take rate, and two-sided interviews — before you build the platform. A practical handbook for B2B, consumer, local, creator, and talent marketplaces.
17 min read
Validation Guide
AI Startup vs SaaS Startup: How Validation Is Different
Why AI startups need workflow, output-quality, and dependency tests on top of every SaaS validation question — and the cheapest experiment that proves each one before you build.
18 min read
Validation Guide
How to Validate an AI Startup Idea
Validate an AI startup idea by testing workflow demand, output quality, pricing, model dependency, distribution, and defensibility before building.
21 min read
Validation Guide
How to Validate a SaaS Idea Before You Build It
The five recurring-revenue assumptions that decide whether a SaaS product survives month six, and the cheapest experiment that tests each one — before you write code.
16 min read
Monetization Validation
How to Test Willingness to Pay Before You Build: Six Experiments, Four Cases, and a Seven-Day Plan
Six pricing experiments you can run this week, four documented startup cases that show the pattern, and a seven-day Stripe-checkout plan that asks for money — before you build.
14 min read
Validation Guide
Startup Validation Checklist: Before You Build
21 concrete checks across four validation stages — problem, customer, business, execution — with how to test each one, the common mistake to avoid, and a printable summary.
17 min read
Validation Guide
How to Validate a Startup Idea Before You Build
The four assumptions every startup depends on, the four questions that test them, and the cheapest experiments that produce evidence in 2–4 weeks — before you build.
17 min read
创始人心理
为什么我们跳过那些能救我们的步骤
Founder Mindset
The Build-vs-Buy Decision for First-Time Founders: When to Ship Software, When to Pay for It, and the Only Test That Picks
A first-time founder's guide to the build-vs-buy decision — when to pay for an existing SaaS, when to assemble a no-code stack, when to fork an open-source project, when to hire a contractor, and when the only honest answer is to build from scratch. The four-question test that picks the right path, the five situations where the answer is pre-decided, and the six-month read-out that proves which side was right.
14 min read
Founder Mindset
How to Position Against an Incumbent: The [for–unlike] Wedge That Holds When a Competitor Already Owns the Category
A founder's guide to writing a positioning wedge when a more-funded, more-distributed competitor already exists in the category — April Dunford's five-component positioning, the four wedge types (segment, persona, capability, distribution), the [for–unlike] sentence template, the five cheap tests that catch a positioning that won't hold, and the failure modes that produce a wedge the team cannot defend in a sales call.
17 min read
Founder Mindset
Why Founders Build Before They Validate
Why the keyboard feels safer than the phone — confirmation bias, action bias, sunk cost, and a small weekly habit that breaks the build-first loop.
15 min read
相关资源
验证涉及创业公司的每一个其他部分。这些资源涵盖了本中心未涉及的部分。
FAQ 中心
常见问题
关于 Startup MRI 如何工作、分数含义以及如何使用报告的直接答案。
了解更多 →词汇表
验证的词汇
每个重要概念的简明定义 — MVP、产品-市场契合度、付费意愿等。
了解更多 →验证工具
创业点子验证工具
创业点子验证工具能做什么、如何在 7 个维度上评分、适用对象，以及免费 Startup MRI 分析如何工作。
了解更多 →评估工具
创业点子评估工具
对你的点子进行 7 维度评估 —— 验证可行性、市场机会、渠道、变现、竞争、构建难度和创始人匹配度。同样的引擎，评估工具的词汇。
了解更多 →评分
创业评分计算器
按 7 个维度给出 0–100 的创业评分，由一句话点子加上一组结构化输入得出。同样的引擎，评分形式的输出。
了解更多 →实操清单
创业验证清单
九个版块、三十多项是/否核对，写代码前就能逐项打勾。中心页的结构化搭档，专为快速浏览设计。
了解更多 →方法
AI 点子验证器的替代
Yibud 对比典型 AI 点子验证器：确定性分数、可选文案润色、免费且无需注册。写给正在比较模型打分工具的创始人。
了解更多 →产品
运行结构化扫描
五个简短问题。60 秒内生成 7 维度结构化报告。报告可免费阅读，其后的执行套件为一次性 $2.99。无需注册，相同输入永远得到相同报告。
了解更多 →