精益画布
- 作者仓库星标 38
- 许可证 MIT
- 作者仓库 builder-skills
- 领域
- 人工智能
- 作者声明的智能体
- 未找到明确声明;不据此推断已兼容或已测试
- 静态检查
- 94 / 100 · 启发式扫描,不代表运行安全
- 运行环境证据
- 未执行;当前没有隔离运行测试证据
- 作者 / 版本 / 许可证
- @kazdenc · MIT
- 上下文用量评级
- 低消耗
- 接入复杂程度
- 需简单配置
- 是否需要外部接口密钥
- 不需要
- 兼容的系统
- 未声明(默认跨平台)
- 运行环境要求
- 未检出明确要求
- 检测到的文件与系统行为
-
- 只读
- 允许写入 / 修改
- 命令执行
- 网络行为
- 仅限本地
- 安装命令数
- 无(仅作为资料)
需要注意: 未声明工具权限范围;实际权限由运行环境决定。
# Lean Canvas
When someone asks you to think through a business model, product strategy, or idea validation, use the lean canvas. Fill one page, not a slide deck. The constraint is the point — it forces clarity.
A lean canvas is a snapshot of a business model hypothesis. Every box is a bet. Treat it that way: state the bet, then figure out how to test it.
## The Canvas Layout
```
┌───────────────┬───────────────┬───────────────┬───────────────┬───────────────┐
│ │ │ │ │ │
│ PROBLEM │ SOLUTION │ UNIQUE │ UNFAIR │ CUSTOMER │
│ │ │ VALUE │ ADVANTAGE │ SEGMENTS │
│ Top 3 │ Top 3 │ PROPOSITION │ │ │
│ problems │ features │ │ Can't be │ Target │
│ │ │ Single, │ easily │ customers │
│ │ │ clear, │ copied or │ │
│ │ │ compelling │ bought │ Early │
│ │ │ message │ │ adopters │
│ │ │ │ │ │
├───────────────┴───────┬───────┴───────────────┴───────┬───────┴───────────────┤
│ │ │ │
│ KEY METRICS │ CHANNELS │ REVENUE STREAMS │
│ │ │ │
│ Key activities │ Path to │ Revenue model │
│ you measure │ customers │ Pricing │
│ │ │ │
├───────────────────────┴───────────────────────────────┴───────────────────────┤
│ │ │
│ COST STRUCTURE │ REVENUE STREAMS │
│ │ │
│ Fixed and variable costs │ (continued from above) │
│ │ │
└───────────────────────────────┴───────────────────────────────────────────────┘
```
Simplified reading order — fill boxes in this sequence for the clearest thinking:
```
1. Customer Segments → 2. Problem → 3. Unique Value Proposition
4. Solution → 5. Channels → 6. Revenue Streams
7. Cost Structure → 8. Key Metrics → 9. Unfair Advantage
```
Start with the customer, not the solution. If someone hands you a solution first, work backward to the customer and problem before filling in anything else.
## The Nine Boxes
### 1. Customer Segments
Put the specific group of people who have the problem. Not "everyone." Not "businesses." Name a narrow early-adopter segment you can reach first.
**Fill with:** One primary segment. Optionally one secondary. Describe them by behavior, not demographics. "Freelance designers who manage their own invoicing" beats "millennials aged 25-35."
**Common mistakes:**
- Too broad ("SMBs," "developers"). Narrow it until you can picture where they hang out online.
- Listing multiple unrelated segments. One canvas = one segment. Make separate canvases for different segments.
- Confusing users with customers (the person who pays may differ from the person who uses).
**Connects to:** Problem (their problem, not yours), Channels (where you find them), Revenue (what they'll pay).
### 2. Problem
List the top 1-3 problems this segment actually has. Not problems you wish they had. Not problems your solution happens to solve.
**Fill with:** Concrete problems stated from the customer's perspective. "I spend 4 hours a week chasing late invoices" not "invoicing is inefficient." Also note existing alternatives — how they solve this problem today (even badly).
**Common mistakes:**
- Listing problems only your solution solves (solution bias). Step back: what's the real problem?
- Skipping existing alternatives. If you don't know how they cope today, you don't understand the problem.
- Too abstract. "Communication is hard" is not a problem — "remote teams lose context when handoffs happen across time zones" is.
**Connects to:** Customer Segments (whose problem), Solution (your answer), UVP (why your answer is better than existing alternatives).
### 3. Unique Value Proposition
Write a single, clear sentence that says why this product deserves attention. This is the hardest box. Don't rush it.
**Fill with:** One sentence. Format: `[End result customer wants] + [Specific time period or qualifier] + [Address the main objection]`. Example: "Get paid on time, every time — without chasing clients."
**Common mistakes:**
- Feature soup ("AI-powered, cloud-native, real-time collaboration platform"). Features go in Solution.
- No differentiation. If you can swap in a competitor's name and it still works, it's not a UVP.
- Too clever. Clarity beats cleverness. The reader should understand it in under 5 seconds.
**Connects to:** Problem (the pain it resolves), Solution (what delivers it), Customer Segments (who cares about this promise).
### 4. Solution
List the top 3 features or capabilities that deliver on the UVP. Keep it minimal — this is your MVP scope.
**Fill with:** Three bullet points max. Each should map directly to a problem from box 2. If a feature doesn't map to a stated problem, cut it.
**Common mistakes:**
- Feature overload. More than 3 means you haven't prioritized.
- Solutions that don't map back to stated problems.
- Building before validating. The solution box is a hypothesis — test it before you commit.
**Connects to:** Problem (each solution maps to a problem), UVP (together they deliver the promise), Cost Structure (what it costs to build and run).
### 5. Channels
Describe how you reach and deliver value to customers. Split into two phases: how they find out about you, and how they get the product.
**Fill with:** Specific channels. "Content marketing" is vague. "SEO-driven blog targeting 'freelance invoice templates'" is specific. Prioritize free or low-cost channels for early stage.
**Common mistakes:**
- Listing aspirational channels you can't actually execute ("viral growth," "partnerships with enterprise").
- Ignoring the channel-segment fit. Your customers are somewhere specific — go there.
- Forgetting delivery channels. How does the product actually reach them?
**Connects to:** Customer Segments (where they are), Cost Structure (channel costs), Revenue Streams (how they buy).
### 6. Revenue Streams
State how you make money and how much you expect to charge. Be specific about the model.
**Fill with:** Pricing model (subscription, one-time, freemium, transaction fee), price point or range, and who pays. Include a rough calculation: `[number of customers] x [price] = [revenue target]`.
**Common mistakes:**
- "We'll figure out pricing later." Price is part of the product. State a hypothesis now.
- No connection to what the customer values. Price should relate to the problem's cost, not your costs.
- Ignoring willingness to pay. What do they pay for existing alternatives?
**Connects to:** Customer Segments (who pays), Cost Structure (margin = revenue minus cost), Channels (how they purchase).
### 7. Cost Structure
List what it costs to operate this business. Include both building the product and reaching customers.
**Fill with:** Fixed costs (salaries, infrastructure, tools) and variable costs (per-customer costs, channel spend). Identify the biggest cost drivers.
**Common mistakes:**
- Only listing product costs, forgetting acquisition costs.
- Being too precise too early. Rough ranges are fine — the point is identifying cost drivers, not budgeting.
- Missing hidden costs (support, compliance, payment processing).
**Connects to:** Revenue Streams (what margin looks like), Solution (build cost), Channels (acquisition cost).
### 8. Key Metrics
Define how you measure whether this model is working. Pick 3-5 metrics max.
**Fill with:** One metric per stage of the customer lifecycle. Use pirate metrics (AARRR) as a scaffold:
| Stage | Measures | Example |
|-------|----------|---------|
| Acquisition | How they find you | Visitors, signups |
| Activation | First "aha" moment | Completed onboarding, first invoice sent |
| Retention | Do they come back | Weekly active use, month-2 retention |
| Revenue | Do they pay | Conversion rate, ARPU |
| Referral | Do they tell others | NPS, referral rate |
**Common mistakes:**
- Vanity metrics (page views, total signups). Measure what predicts business health.
- Too many metrics. If everything is a key metric, nothing is.
- No leading indicators. Revenue is a lagging indicator — track what predicts it.
**Connects to:** All other boxes. Metrics validate every hypothesis on the canvas.
### 9. Unfair Advantage
Name something you have that can't be easily copied or bought. This is the moat. It's okay to leave this blank early — most startups don't have one yet.
**Fill with:** Insider knowledge, existing audience, network effects, community, proprietary data, personal authority, regulatory advantage, or team expertise. Not features — features can be copied.
**Common mistakes:**
- Listing things that aren't actually hard to copy ("first mover advantage," "great UX," "passion").
- Confusing competitive advantage with unfair advantage. A competitive advantage is temporary; an unfair advantage compounds over time.
- Forcing it. If you don't have one, write "None yet — to be developed" and plan how to build one.
**Connects to:** UVP (amplifies it), Customer Segments (why they stay), Key Metrics (retention and referral).
## When to Use Lean Canvas vs Other Frameworks
| Situation | Use | Why |
|-----------|-----|-----|
| Validating a new business idea | **Lean Canvas** | Maps the full business model in one page |
| Understanding what customers need | **JTBD** | Goes deeper on the job and unmet needs |
| Prioritizing features in a backlog | **JTBD opportunity scoring** | Quantifies importance vs satisfaction |
| Pitching to investors | Lean Canvas as starting point, then expand | Forces you to articulate the core model |
| Evaluating a pivot | **Lean Canvas** side by side | Compare current model vs proposed model |
| Designing a specific feature | **JTBD job stories** | "When [situation], I want to [job], so I can [need]" |
| Competitive analysis | Both | Lean Canvas maps their model; JTBD maps how well they satisfy needs |
| Planning a product launch | Neither — use a launch plan | Lean Canvas is strategy, not execution |
Use lean canvas for breadth (whole business model). Use JTBD for depth (customer needs and jobs). They're complementary.
## Lean Canvas to JTBD Mapping
When both frameworks are relevant, connect them:
| Lean Canvas Box | JTBD Concept | How They Connect |
|-----------------|-------------|------------------|
| **Customer Segments** | Job performer | Same person — described by the job they're hiring a product to do |
| **Problem** | Underserved needs | Problems are clusters of unmet needs within a job |
| **Existing Alternatives** | Current solutions being "fired" | The push force — what's driving switching behavior |
| **UVP** | Why they'd "hire" your product | The pull force — the promise that attracts job performers |
| **Solution** | Product capabilities | Features that satisfy the most underserved needs |
| **Customer Segments (early adopters)** | Most-constrained performers | People whose needs are most underserved by current solutions |
| **Key Metrics (Activation)** | Job completion | The moment the customer successfully gets the job done |
| **Key Metrics (Retention)** | Recurring job frequency | How often the job occurs determines natural usage cadence |
| **Revenue Streams** | Willingness to pay | Correlates with importance of the job and pain of current alternatives |
| **Unfair Advantage** | Switching costs (in your favor) | What creates anxiety about leaving — network effects, data, habit |
| **Channels** | Circumstances / context | Where and when customers encounter the need to get the job done |
## Working With the Canvas
**Fill it in 20 minutes, not 20 days.** Speed matters. The first version is wrong — that's the point. You're making your assumptions explicit so you can test them.
**One canvas per customer segment.** If you have two distinct segments, make two canvases. Merging them hides critical differences.
**Highlight the riskiest box.** After filling it in, circle the box you're least confident about. That's what to test first. Usually it's Problem or Customer Segments, not Solution.
**Revise weekly during early stage.** The canvas is a living document. After each customer conversation, experiment, or data point, update it.
**Read it as a story.** A good canvas tells a coherent narrative: "[Customer] has [problem]. They currently [existing alternative]. We offer [UVP] through [solution], reaching them via [channels]. We charge [revenue model] and it costs us [cost structure]. We'll know it's working when [key metrics]. Over time, [unfair advantage] makes us hard to displace."
**Use it to pressure-test ideas people bring you.** When someone says "we should build X," put it on a canvas. The empty boxes reveal what they haven't thought through.
章节与文件索引
按说明文件中的顺序排列。
画布布局
1. 客户群
2. 问题
3. 独特的价值主张
4、解决方案
文件/命令 -> [End result customer wants] + [Specific time period or qualifier] + [Address the main objection] · [number of customers] x [price] = [revenue target]
内容校验值 -> 741a7f173a05
当前原文的自动摘译,尚未人工审校。
画布布局
简化的阅读顺序 - 按此顺序填写方框以获得最清晰的思维:从客户开始,而不是解决方案。如果有人先给你一个解决方案,请先回溯到客户和问题,然后再填写其他内容。
1. 客户群
列出有问题的特定人群。不是“每个人”。不是“企业”。说出您可以首先接触到的一个狭窄的早期采用者群体。填充:一个主要部分。可选一二级。通过行为而不是人口统计来描述他们。 “自己管理发票的自由设计师”击败了“25-35 岁的千禧一代”。常见错误:太宽泛(“中小企业”、“开发人员”)。缩小范围,直到你能想象出他们在网上闲逛的地方。
2. 问题
列出该细分市场实际存在的前 1-3 个问题。不是您希望他们遇到的问题。不是你的解决方案恰好解决的问题。填写:从客户角度陈述的具体问题。 “我每周花 4 个小时追讨迟到的发票”,而不是“开具发票效率低下”。还要注意现有的替代方案——它们今天如何解决这个问题(甚至很糟糕)。常见错误:列出只有您的解决方案才能解决的问题(解决方案偏差)。退一步:真正的问题是什么?
3. 独特的价值主张
写一个简单、清晰的句子来说明为什么该产品值得关注。这是最难的盒子。别着急。填:一句话。格式:
[End result customer wants] + [Specific time period or qualifier] + [Address the main objection]。示例:“每次都按时获得报酬,无需追逐客户。”常见错误:功能汤(“人工智能驱动、云原生、实时协作平台”)。功能包含在解决方案中。4、解决方案
列出 UVP 提供的前 3 个特性或功能。保持最小化——这是你的 MVP 范围。填写:最多三个要点。每个功能都应直接映射到框 2 中的问题。如果某个功能未映射到指定的问题,则将其删除。常见错误:功能过载。超过 3 意味着您尚未确定优先顺序。解决方案不映射回所述问题。