ISOM5320 Quiz Review

今晚 90 分钟复习路线

0-20 min先背框架
平台经济、生态控制点、护城河、IP、创作者平台、FinTech、AI extraction、subscription triangle。Memorize the core frameworks first: platform economics, ecosystem control points, moats, IP, creator platforms, FinTech, AI extraction, and the subscription triangle.

20-55 min看案例对比
Apple vs Nokia, Netflix vs Disney, Meituan vs JD/Alibaba, Sanrio vs Pop Mart, Ant/JD/Waterdrop, iCloud/Netflix/Eight Sleep/ChatGPT。Review the case comparisons: Apple vs Nokia, Netflix vs Disney, Meituan vs JD/Alibaba, Sanrio vs Pop Mart, Ant/JD/Waterdrop, and iCloud/Netflix/Eight Sleep/ChatGPT.

55-80 min做 MCQ
选择题重点通常是概念辨析：platform vs pipeline, content vs IP, KOL vs KOC, active vs passive subscription。Practice MCQs. They usually test concept distinctions: platform vs pipeline, content vs IP, KOL vs KOC, and active vs passive subscription.

80-90 min背简答模板
每题按“定义 → 机制 → 案例 → 风险/局限 → 结论”写，英文名词尽量写出来。Memorize the short-answer template: definition → mechanism → case → risk/limitation → conclusion. Use English terms whenever possible.

高频术语中英双语解释

Platform 平台连接多边用户并促进互动的商业架构；价值来自互动，不只是内部生产。A business architecture that connects multiple sides and enables interactions; value comes from interactions, not only internal production.

Pipeline 管道型企业线性价值链：input → process → output，例如传统制造、零售。扩张通常更线性。A linear value chain: input → process → output, such as traditional manufacturing or retail. Scaling is usually more linear.

Network Effects 网络效应用户越多，网络对每个用户越有价值。Metcalfe's Law 认为价值大致随 n² 增长。The more users a network has, the more valuable it becomes to each user. Metcalfe's Law suggests value roughly grows with n².

Zero Marginal Cost 零边际成本数字产品一旦建好，复制给额外用户的成本接近 0。Once a digital product is built, serving one more user costs almost nothing.

Flywheel 飞轮多个环节互相强化形成增长循环，例如 Amazon 的低价、更多用户、更多卖家、更好选择。A reinforcing growth loop where multiple parts strengthen each other, such as Amazon's lower prices, more customers, more sellers, and better selection.

Ecosystem Strategy 生态战略通过互补伙伴、开发者、商家、用户共同创造价值，而非只靠单个产品。A strategy that creates value through complementors, developers, merchants, and users, rather than relying on a single product.

Control Point 控制点生态中最难替代、最能控制价值流的位置，如支付、物流、身份、算法分发、商家 SaaS。A hard-to-replace position in an ecosystem that controls value flows, such as payment, logistics, identity, algorithmic distribution, or merchant SaaS.

Moat 护城河防止竞争者复制或夺走利润的结构性优势；课程强调运营护城河常比纯技术更难复制。A structural advantage that prevents competitors from copying the business or taking profits; the course emphasizes that operational moats often beat pure technology moats.

Transient Advantage 暂时性优势数字时代优势生命周期缩短，企业要不断创造、培育、利用、退出和重配优势。In the digital era, advantages have shorter lifespans, so firms must continuously create, nurture, exploit, abandon, and reallocate advantages.

IP 知识产权/角色宇宙可跨品类、跨代际延展的角色、世界观或宇宙；IP 会增值，普通内容会折旧。A character, world, or universe that can extend across categories and generations; IP appreciates while ordinary content depreciates.

Creator Economy 创作者经济个人或团队通过平台分发内容、积累影响力并商业化，包括广告、品牌、直播电商、带货、会员等。An economy where individuals or teams distribute content through platforms, build influence, and monetize through ads, brands, live commerce, affiliate sales, or membership.

KOL / KOCKOL 是关键意见领袖，重影响力和权威；KOC 是关键意见消费者，重真实感和同伴信任。KOL means Key Opinion Leader, emphasizing reach and authority; KOC means Key Opinion Consumer, emphasizing authenticity and peer trust.

Financial Inclusion 普惠金融让原本无法获得银行服务的人群或 SME 获得支付、信贷、保险、理财等金融服务。Providing payment, credit, insurance, and wealth services to people or SMEs previously excluded from formal banking.

Alternative Data 替代数据非传统征信数据，如交易记录、支付行为、物流、浏览、经营数据，用于风险评估。Non-traditional credit data, such as transactions, payment behavior, logistics, browsing, and business operations, used for risk assessment.

Knowledge Extractivism 知识攫取主义企业一次性获取个人或社群的具身知识，再用 AI 永久复制或替代来源，并捕获未来剩余价值。A strategy where a firm acquires embodied knowledge once, uses AI to replicate or replace its source permanently, and captures future surplus.

Digital Rent 数字租来自控制稀缺数字资源的收益，如平台入口、数据、算法模型，而不完全是新创造价值。Income earned by controlling scarce digital resources, such as platform access, data, or algorithmic models, rather than purely creating new value.

Anti-Commons 反公地悲剧资源被过度私有化后反而生产不足；AI 抓取可能导致创作者停止公开发布，未来训练数据减少。A situation where over-privatization leads to under-production; in AI, scraping may cause creators to stop publishing publicly, reducing future training data.

Subscription Triangle 订阅三角判断订阅是否健康的三问：价值是否持续、边际成本如何、用户关系是什么。A three-question test for subscriptions: whether value is continuous, how marginal cost behaves, and what relationship exists between user and firm.

Week 1-8 Cheat Sheet

如果你觉得概念很散，先看这一段。课程主线其实是：数字企业为什么容易变成平台 → 平台怎么变成生态 → 生态如何形成护城河 → 内容/IP/创作者/金融/订阅/AI 如何成为新的增长曲线。If the concepts feel scattered, start here. The course storyline is: why digital firms become platforms → how platforms become ecosystems → how ecosystems build moats → how content/IP/creators/finance/subscriptions/AI become new growth curves.

阅读方法：中文帮你理解，英文帮你考试表达。How to read this: use the Chinese for understanding and the English for exam wording.

先看懂每个 PDF 在讲什么

Week1_Digital Business.pdf数字商业的底层规则
这份是全课入口：为什么互联网公司和传统公司不一样。重点不是某家公司，而是几个基础规律：平台比管道更容易放大、用户越多越有价值、软件复制成本很低、中国互联网常用免费和补贴先抢规模。This is the entry point of the course: why internet firms differ from traditional firms. The focus is not one company, but the basic rules: platforms scale better than pipelines, more users create more value, software has near-zero replication cost, and Chinese internet firms often use free products and subsidies to capture scale first.
你要抓住 / Key terms: platform, network effects, zero marginal cost, flywheel, disruptive innovation。

Week 2_Platforms, Ecosystems, and Beyond.pdf平台如何升级成生态
这一份讲“有用户”还不够，强公司会把支付、身份、算法、物流、商家工具等变成生态控制点。腾讯、阿里、淘宝、微信、抖音都是不同类型的生态。This PDF explains why “having users” is not enough. Strong platforms turn payment, identity, algorithms, logistics, and merchant tools into ecosystem control points. Tencent, Alibaba, Taobao, WeChat, and TikTok represent different ecosystem types.
你要抓住 / Key terms: ecosystem strategy, control points, WeChat Pay vs Alipay, transient advantage。

Week3.1 Ecosystem Building.pdfNetflix 为什么能一直活下来
这份基本是 Netflix 案例故事。它讲的不是“Netflix 很会拍剧”，而是 Netflix 会主动破坏自己旧业务：DVD、流媒体、原创内容、广告、游戏/直播，每次都比别人更早转向。This is mainly the Netflix case. The point is not simply that Netflix makes good shows, but that Netflix repeatedly self-disrupts: DVD, streaming, original content, advertising, games/live events. It pivots before competitors force it to.
你要抓住 / Key terms: self-disruption, innovator's dilemma, data-driven content, content platform flywheel。

Week3.2 Ecosystems and Moat.pdf护城河到底是什么
这一份用 JD 物流、美团外卖/即时零售大战说明：真正难复制的不是一个 App 功能，而是长期堆出来的履约网络、商家关系、用户习惯和运营密度。补贴可以买短期订单，但买不到真正的 moat。This PDF uses JD Logistics and the Meituan food-delivery/instant-retail war to explain moats. What is hard to copy is not an app feature, but fulfillment networks, merchant relationships, user habits, and operational density built over years. Subsidies can buy temporary orders, but not a real moat.
你要抓住 / Key terms: operational moat, fulfillment density, nominal market vs real market, four-moat test。

Week4.1 IP.pdf为什么 Disney/Hello Kitty/Pokémon 比普通内容更值钱
这一份讲 IP 生意。普通 content 看完就折旧，IP 会跨电影、游戏、乐园、商品、授权继续赚钱。Disney 靠故事宇宙，Sanrio 靠情绪投射，Nintendo/Pokémon 靠游戏体验，Pop Mart 靠收集和盲盒。This PDF is about IP businesses. Ordinary content depreciates after consumption, while IP can keep earning across films, games, theme parks, merchandise, and licensing. Disney relies on narrative universes, Sanrio on emotional projection, Nintendo/Pokémon on game experiences, and Pop Mart on collecting and blind boxes.
你要抓住 / Key terms: content vs brand vs IP, IP flywheel, licensing, scarcity, emotional attachment。

Week4.2 Creator Economy.pdf为什么同一个创作者换平台就赚不到钱
这一份讲创作者经济。核心不是“谁粉丝多”，而是平台收入机制决定创作者命运：YouTube 适合长视频和广告分成，TikTok 适合注意力和电商转化，小红书适合种草和品牌合作。This PDF is about the creator economy. The core question is not only who has more followers, but how platform economics shapes creator income: YouTube rewards long-form video and ad sharing, TikTok rewards attention and e-commerce conversion, and Xiaohongshu rewards product seeding and brand deals.
你要抓住 / Key terms: platform economics, KOL/KOC, MrBeast, Kardashian, Li Jiaqi, trust-based selling。

Week5.1 Fintech model.pdf平台为什么会做金融
这一份讲平台成熟后怎么找第二增长曲线：支付先建立信任和数据，然后进入信贷、理财、保险。Ant、JD Finance、Waterdrop 都是在用平台数据解决传统金融服务不了的长尾用户/SME。This PDF explains why mature platforms move into finance as a second growth curve. Payment creates trust and data, then expands into credit, wealth management, and insurance. Ant, JD Finance, and Waterdrop all use platform data to serve long-tail users or SMEs that traditional finance fails to serve.
你要抓住 / Key terms: financial inclusion, alternative data, cost-risk-scale triangle, CreditTech, regulation。

Lesson7 Knowledge Extraction.pdfAI 时代谁创造价值、谁拿走价值
这一份比较理论，但核心很简单：公司一次性买走或抓取人的经验、数据、作品，用 AI 永久复制，然后未来收益主要归公司。传统经济学很多概念解释不了这种结构。This lesson is theoretical, but the core idea is simple: firms buy or scrape human experience, data, and creative work once, use AI to replicate it permanently, and capture most future surplus. Many traditional economic theories struggle to explain this structure.
你要抓住 / Key terms: knowledge extractivism, primitive accumulation, digital rent, anti-commons, human capital becomes liability。

Lesson8 Subscription Models.pdf什么订阅是真护城河，什么只是圈钱
这一份教你判断订阅模式：用户停付会失去什么？重度用户是赚钱还是亏钱？用户是客户还是被卖给广告主的产品？iCloud、Netflix、Eight Sleep、ChatGPT、Xiaomi TV 是五种不同答案。This lesson teaches how to judge subscription models: what does the user lose if they stop paying? Are heavy users profitable or loss-making? Is the user the customer, or the product sold to advertisers? iCloud, Netflix, Eight Sleep, ChatGPT, and Xiaomi TV give five different answers.
你要抓住 / Key terms: subscription triangle, value continuity, marginal cost, co-creative/hostage/extractive relationship。

每周核心概念速记

Week 1 Digital Business Foundation

核心逻辑：数字商业有网络效应、零边际成本、数据反馈循环，因此容易出现 winner-takes-all。Core logic: digital business has network effects, near-zero marginal cost, and data feedback loops, so it often tends toward winner-takes-all outcomes.
Platform vs Pipeline：平台连接多边市场，规模可指数增长；管道依赖内部生产，规模更线性。Platform vs pipeline: platforms connect multiple market sides and can scale exponentially; pipelines rely on internal production and scale more linearly.
中国互联网逻辑：free-first + subsidy-to-scale + burn → scale → dominate，先抢用户和数据，再商业化。Chinese internet logic: free-first plus subsidy-to-scale; firms burn cash to gain scale, capture users/data, and monetize later.
Apple vs Nokia：Apple 是 iOS ecosystem，有开发者、App Store、iCloud、iMessage 与高 switching cost；Nokia 更像硬件 pipeline。Apple vs Nokia: Apple built an iOS ecosystem with developers, App Store, iCloud, iMessage, and high switching costs; Nokia remained closer to a hardware pipeline.
Disruptive innovation：新进入者先服务低端或新市场，再上移替代 incumbent。TikTok 用短视频、算法分发和创作者平台冲击传统媒体。Disruptive innovation: entrants first serve low-end or new markets, then move upmarket to displace incumbents. TikTok disrupted traditional media through short video, algorithmic distribution, and creator participation.

Week 2 Platforms, Ecosystems, and Beyond

Ecosystem strategy：长期价值不只在产品，而在互补者和控制点组成的生态。Ecosystem strategy: long-term value lies not only in products, but in the ecosystem of complementors and control points.
生态类型：Taobao platform-led, WeChat orchestrator-led, TikTok algorithm-driven。Ecosystem types: Taobao is platform-led, WeChat is orchestrator-led, and TikTok is algorithm-driven.
WeChat Pay vs Alipay：支付是生态基础设施，红包和社交关系让 WeChat Pay 快速突破。WeChat Pay vs Alipay: payment is ecosystem infrastructure; red packets and social ties helped WeChat Pay break through quickly.
Operational moat：JD 物流、美团骑手网络和商家 SaaS、滴滴调度算法都比单纯技术更难复制。Operational moat: JD logistics, Meituan's rider network and merchant SaaS, and Didi's dispatch algorithms are harder to copy than pure technology features.
Transient advantage：不要只守一个 moat，要持续创新、测试、扩张、退出并重配资源。Transient advantage: firms should not defend one moat forever; they must continuously innovate, test, scale, exit, and reallocate resources.

Week 3.1 Netflix Case

Netflix 六次自我颠覆：DVD no late fee → streaming → original content → global content → ads/paid sharing → entertainment ecosystem。Netflix self-disrupted six times: DVD/no late fees → streaming → original content → global content → ads/paid sharing → entertainment ecosystem.
Blockbuster 输在 innovator's dilemma：late fees 是利润池，不能自我破坏；Netflix 正好攻击其利润中心。Blockbuster lost because of the innovator's dilemma: late fees were its profit pool, so it could not destroy them; Netflix attacked exactly that profit center.
House of Cards：data-driven content bet，不看 pilot，基于用户偏好、演员、导演重叠数据下注。House of Cards was a data-driven content bet: Netflix committed without a pilot, based on overlapping viewer preferences for the cast, director, and story type.
Netflix moat：data network effects, global content breadth, scale amortization, self-cannibalization capability。Netflix's moat: data network effects, global content breadth, scale amortization, and the organizational ability to self-cannibalize.
2022 危机后 pivot：广告 tier 和 password sharing crackdown，把 freeloaders 转成收入。After the 2022 crisis, Netflix pivoted to ad tiers and password-sharing crackdowns, converting freeloaders into revenue.

Week 3.2 Ecosystems and Moat

JD logistics：控制仓储、履约、last-mile，形成高固定成本但高防御性的 operational moat。JD Logistics controls warehousing, fulfillment, and last-mile delivery, creating a high-fixed-cost but defensible operational moat.
Meituan：核心资产是 last-mile network + merchant SaaS；外卖只是即时零售 fulfillment layer 的入口。Meituan's core assets are its last-mile network and merchant SaaS; food delivery is the entry point to the broader instant-retail fulfillment layer.
Food delivery war 2025：JD 和 Alibaba 用补贴买 nominal market share；Meituan 不全面跟价格战，保护真实护城河。In the 2025 food-delivery war, JD and Alibaba bought nominal market share through subsidies; Meituan avoided a full price war to protect its real moat.
四护城河测试：fulfillment density, merchant lock-in, user habit durability, cross-subsidy sustainability。Four-moat test: fulfillment density, merchant lock-in, user habit durability, and cross-subsidy sustainability.
Nominal vs real market：补贴买来的订单不等于真实市场，补贴结束后留下来的用户、骑手、商家才算 moat。Nominal vs real market: orders bought by subsidies are not the real market; the users, riders, and merchants who remain after subsidies end are the real moat.

Week 4.1 IP-Driven Digital Business

Content 内容：被消费后折旧；Brand 品牌：创造偏好但延展有限；IP：角色/世界观可跨品类、跨代际增值。Content depreciates after consumption; brands create preference but have limited extensibility; IP can appreciate across categories and generations.
Disney：IP flywheel = create IP → stream → parks/experience → merchandise/license → reinvest。Disney's IP flywheel: create IP → stream it → turn it into parks/experiences → monetize through merchandise/licensing → reinvest.
Sanrio：Hello Kitty 是 emotional blank canvas，无固定叙事，极易授权，royalty margin 高。Sanrio: Hello Kitty is an emotional blank canvas with no fixed narrative, making it highly licensable with high royalty margins.
Nintendo/Pokémon：hardware is delivery mechanism for IP experiences；独占和控制稀缺防止疲劳。Nintendo/Pokémon: hardware is the delivery mechanism for IP experiences; exclusivity and controlled scarcity prevent franchise fatigue.
Pop Mart：collectibility + scarcity + blind box + community，但有单一 IP 依赖和监管风险。Pop Mart: collectibility, scarcity, blind boxes, and community drive growth, but single-IP dependence and regulatory risk remain concerns.

Week 4.2 Creator Economy

总 thesis：Platforms shape creators; culture shapes trust; infrastructure shapes monetization。Main thesis: platforms shape creators; culture shapes trust; infrastructure shapes monetization.
YouTube：media + subscription company，55% ad revenue share，long-form 可支撑 MrBeast 高成本内容。YouTube is a media and subscription company. Its 55% ad revenue share and long-form format can support MrBeast's high-cost production model.
TikTok：business of attention + Shop loop，creator often becomes merchandise sold to advertisers。TikTok is a business of attention plus a Shop loop; creators often become the merchandise sold to advertisers.
Xiaohongshu：purchase decision search engine，平台不直接分成给创作者，KOL/KOC 靠品牌合作和种草赚钱。Xiaohongshu is a purchase-decision search engine. The platform pays creators little directly; KOLs/KOCs earn through brand deals and product seeding.
三类创作者商业化：MrBeast content → brand；Kardashian attention → equity；Li Jiaqi trust → sales。Three creator monetization models: MrBeast turns content into brands; Kardashian turns attention into equity; Li Jiaqi turns trust into sales.
AI 影响：commodity content 更易被替代；trust, identity, authenticity 更抗 AI。AI impact: commodity content is easier to replace; trust, identity, and authenticity are more AI-resistant.

Week 5 FinTech Model

传统金融排斥穷人/SME 的原因：high unit cost, perceived high risk, physical infrastructure limits。Traditional finance excludes the poor and SMEs because of high unit cost, perceived high risk, and physical infrastructure limits.
FinTech 改写规则：mobile onboarding, remote KYC, instant approval, alternative data, near-zero marginal cost。FinTech rewrites the rules through mobile onboarding, remote KYC, instant approval, alternative data, and near-zero marginal cost distribution.
Ant Group：Alipay escrow 建立信任与数据，扩展到 Yu'e Bao, Huabei, Jiebei, insurance, wealth。Ant Group: Alipay escrow built trust and data, then expanded into Yu'e Bao, Huabei, Jiebei, insurance, and wealth management.
Iron Triangle of Finance：cost, risk, scale 三者传统上难兼得；FinTech 通过数据和平台分发降低成本、改善风险评估。Iron Triangle of Finance: cost, risk, and scale are traditionally hard to achieve together; FinTech lowers cost and improves risk assessment through data and platform distribution.
监管担忧：shadow banking, capital arbitrage, consumer protection, systemic risk。Regulatory concerns: shadow banking, capital arbitrage, consumer protection, and systemic risk.
Waterdrop：mutual aid 是用户获取 funnel，不是利润核心；问题是 adverse selection、成本螺旋和监管不确定。Waterdrop: mutual aid was a customer-acquisition funnel, not the core profit engine; problems included adverse selection, cost spirals, and regulatory uncertainty.
JD Finance vs GLP：GLP 是 physical assets → financial instruments；JD 是 digital data → credit assets。JD Finance vs GLP: GLP converts physical assets into financial instruments; JD converts digital data into credit assets.

Week 7 The Great Knowledge Extraction

Knowledge extractivism：一次性买/抓知识，用 AI 永久复制，价值创造者和价值捕获者分离。Knowledge extractivism: firms buy or scrape knowledge once, use AI to replicate it permanently, and separate value creation from value capture.
四种形式：talent buyout, dataset acquisition, skill distillation by observation, creative capture from open web。Four forms: talent buyout, dataset acquisition, skill distillation by observation, and creative capture from the open web.
五个理论危机：marginal pricing → permanent replacement; Coase property rights → strategic ambiguity; Becker human capital → human liability; Schumpeter creative destruction → destructive creation; Akerlof information asymmetry → information evaporation。Five theoretical crises: marginal pricing becomes permanent replacement; Coasean property rights become strategic ambiguity; Becker's human capital becomes human liability; Schumpeter's creative destruction becomes destructive creation; Akerlof's information asymmetry becomes information evaporation.
三个理论工具：primitive accumulation, digital rent, anti-commons。Three theoretical tools: primitive accumulation, digital rent, and anti-commons.
策略路径：public-good data pool, progressive open-sourcing, experience royalties, data provenance labels, data trusts/unions。Strategy paths: public-good data pools, progressive open-sourcing, experience royalties, data provenance labels, and data trusts/unions.

Week 8 Subscription Business Models

Subscription Triangle：value continuity, marginal cost structure, relationship architecture。Subscription Triangle: value continuity, marginal cost structure, and relationship architecture.
iCloud/Salesforce：accumulating value + decreasing cost + co-creative relation = holy grail。iCloud/Salesforce: accumulating value plus decreasing cost plus co-creative relationship equals the holy grail.
Netflix：refreshing value + flat/library economics + servicing relation = healthy but exposed。Netflix: refreshing value plus flat/library economics plus a servicing relationship makes it healthy but exposed.
Eight Sleep：static value + hostage relation = ransom/pseudo-subscription。Eight Sleep: static value plus a hostage relationship makes it a ransom-like pseudo-subscription.
ChatGPT Plus：refreshing value but increasing marginal cost；fixed price + unlimited usage + latest model 三者只能选二。ChatGPT Plus: refreshing value but increasing marginal cost; fixed price, unlimited usage, and latest model access cannot all hold at once.
Xiaomi TV：passive subscription，用户不付订阅费，但通过广告注意力和数据“支付”。Xiaomi TV: passive subscription; users do not pay a monthly fee, but they “pay” through advertising attention and data.

简答题万能框架

推荐结构：Define 概念定义 → Mechanism 机制 → Case 课程案例 → Tension 风险/局限 → Conclusion 回答题目。Recommended structure: Define the concept → explain the mechanism → use a course case → discuss tension/limitation → answer the question directly.

Platform answer
先区分 platform/pipeline，再写 network effects、zero marginal cost、data feedback、control point，最后放 Apple/Meituan/TikTok。First distinguish platform from pipeline, then discuss network effects, zero marginal cost, data feedback, and control points. Use Apple, Meituan, or TikTok as the case.

Moat answer
不要只写“技术强”。写 operational moat、switching cost、data advantage、ecosystem lock-in，再比较 nominal vs real market。Do not just say “strong technology.” Discuss operational moat, switching cost, data advantage, ecosystem lock-in, and the difference between nominal and real market share.

IP answer
先区分 content/brand/IP，再用 Disney/Sanrio/Nintendo/Pop Mart 说明 IP 如何跨渠道变现和保持稀缺。First distinguish content, brand, and IP. Then use Disney, Sanrio, Nintendo, or Pop Mart to explain how IP monetizes across channels and preserves scarcity.

Creator answer
用三层框架：platform economics + cultural context + infrastructure。平台不是中性舞台，而是会塑造创作者类型。Use the three-layer framework: platform economics, cultural context, and infrastructure. Platforms are not neutral stages; they shape what type of creator can succeed.

FinTech answer
用 cost-risk-scale 或 traditional banking vs digital rails，对应 Ant/JD/Waterdrop 案例，再补监管 tension。Use the cost-risk-scale framework or compare traditional banking with digital rails. Connect the answer to Ant, JD Finance, or Waterdrop, and add regulatory tension.

Subscription answer
三问：停付损失什么？第 1000 个重度用户赚钱还是亏钱？用户是 customer 还是 product？Ask three questions: what does the user lose if they stop paying? Is the 1,000th heavy user profitable or loss-making? Is the user the customer or the product?

选择题练习 MCQ

先自己做，再点开答案。考试如果问“best answer”，通常选最能体现课程框架的选项。Try each question first, then open the answer. If the exam asks for the “best answer,” choose the option that best reflects the course framework.

1. Apple 相比 Nokia 的核心优势最符合哪个概念？What concept best explains Apple's core advantage over Nokia?

A. Lower manufacturing costB. Platform ecosystem and switching costsC. Offline distributionD. Price war

查看答案 / Show answer

答案 / Answer: B. Apple 有 iOS、App Store、developer network、iCloud/iMessage。Apple built an ecosystem with iOS, App Store, developers, iCloud, and iMessage.

2. “每多一个用户，网络整体价值不只是增加而是放大”对应什么？Which concept means that each additional user multiplies the value of the network?

A. Economies of scopeB. Network effects / Metcalfe's LawC. Adverse selectionD. Licensing

查看答案 / Show answer

答案 / Answer: B.

3. 美团对 JD/Alibaba 补贴战不全面跟进，主要是为了保护什么？Why did Meituan avoid fully matching JD/Alibaba's subsidy war?

A. Nominal shareB. Fulfillment moat and margin structureC. App rankingD. Brand logo

查看答案 / Show answer

答案 / Answer: B. 课程强调让出短期名义份额，保护真实履约网络。The course emphasizes sacrificing short-term nominal share to protect the real fulfillment network.

4. Netflix 早期攻击 Blockbuster 的哪一个利润中心？Which Blockbuster profit center did Netflix attack early on?

A. Popcorn salesB. Late feesC. Theme parksD. Live sports

查看答案 / Show answer

答案 / Answer: B. No late fee 直接打击 Blockbuster 的 late fee revenue。The no-late-fee model directly attacked Blockbuster's late-fee revenue.

5. House of Cards 案例主要说明 Netflix 的什么能力？What capability does the House of Cards case mainly illustrate?

A. Pure luckB. Data-driven original content decisionC. Hardware exclusivityD. Blind-box design

查看答案 / Show answer

答案 / Answer: B.

6. “Content depreciates, IP appreciates” 最适合解释哪组对比？Which comparison best fits “content depreciates, IP appreciates”?

A. Disney vs NetflixB. JD vs SF ExpressC. Ant vs banksD. YouTube vs TikTok

查看答案 / Show answer

答案 / Answer: A.

7. Hello Kitty 没有固定表情和故事，商业上反而有利，因为它是：Why is Hello Kitty's lack of fixed expression and story commercially useful?

A. Emotional blank canvasB. Logistics moatC. Search engineD. Credit bureau

查看答案 / Show answer

答案 / Answer: A. 消费者可投射自己的情绪，授权边界更宽。Consumers can project their own emotions onto it, making licensing more flexible.

8. Pop Mart blind box 的商业机制核心是：What is the core commercial mechanism of Pop Mart's blind boxes?

A. Subscription lock-inB. Collectibility, scarcity, variable rewardC. Remote KYCD. Public data pool

查看答案 / Show answer

答案 / Answer: B.

9. YouTube 为什么更适合 MrBeast 高成本视频？Why is YouTube better suited for MrBeast's high-cost videos?

A. 只做短视频 / Only short videosB. 55% ad share, long-form ads, evergreen discoveryC. 没有广告 / No adsD. 没有订阅用户 / No subscribers

查看答案 / Show answer

答案 / Answer: B.

10. Xiaohongshu 的创作者收入主要来自：Where does creator income on Xiaohongshu mainly come from?

A. Platform direct payB. Brand deals and seedingC. Bank interestD. Hardware margin

查看答案 / Show answer

答案 / Answer: B. 平台直接分成几乎为 0。Direct platform revenue sharing is close to zero.

11. KOC 相比 KOL 的优势通常是：What is the usual advantage of KOC over KOL?

A. 更像明星 / More celebrity-likeB. 更高真实性和同伴信任 / Higher authenticity and peer trustC. 更强硬件能力 / Stronger hardware capabilityD. 更高监管资本 / Higher regulatory capital

查看答案 / Show answer

答案 / Answer: B.

12. 传统银行难服务小额贷款，最直接原因之一是：One direct reason traditional banks struggle with small loans is:

A. Small loans have high fixed processing cost relative to sizeB. 用户太多 / Too many usersC. 手机太便宜 / Phones are too cheapD. IP 太强 / IP is too strong

查看答案 / Show answer

答案 / Answer: A.

13. Ant 从 Alipay 扩张到 Huabei/Jiebei 的关键资源是：What key resource allowed Ant to expand from Alipay into Huabei/Jiebei?

A. Theme parksB. Transaction data and embedded payment trustC. Blind boxesD. DVD mailing

查看答案 / Show answer

答案 / Answer: B.

14. Waterdrop Mutual Aid 的 mutual aid 主要功能是：What was the main function of Waterdrop Mutual Aid?

A. Core profit engineB. Customer acquisition funnelC. Hardware salesD. Theme park experience

查看答案 / Show answer

答案 / Answer: B.

15. Knowledge extractivism 的核心问题是：What is the core problem of knowledge extractivism?

A. 知识没有价值 / Knowledge has no valueB. One-time transfer enables permanent AI replication and surplus captureC. 用户太少 / Too few usersD. 物流太慢 / Logistics are too slow

查看答案 / Show answer

答案 / Answer: B.

16. “AI 训练后无法 untrain，类似圈地运动”对应哪个理论工具？Which theory fits the idea that AI models cannot be “untrained,” similar to enclosure?

A. Primitive accumulationB. KOCC. Subscription triangleD. KYC

查看答案 / Show answer

答案 / Answer: A.

17. Anti-commons 在 AI 语境下的风险是：What is the AI-era risk of anti-commons?

A. 数据越抓越多所以永远可持续 / More scraping makes it sustainable foreverB. 创作者停止公开发布，未来训练数据减少 / Creators stop publishing publicly, reducing future training dataC. 物流成本降为 0 / Logistics cost falls to zeroD. 用户无法取消订阅 / Users cannot cancel subscriptions

查看答案 / Show answer

答案 / Answer: B.

18. iCloud 被视为订阅 holy grail，因为：Why is iCloud considered a subscription holy grail?

A. Static value + hostageB. Accumulating value + decreasing cost + co-creative relationC. Increasing GPU costD. Blind box scarcity

查看答案 / Show answer

答案 / Answer: B.

19. Eight Sleep 的问题更接近：Eight Sleep's problem is closest to:

A. Ransom / hostage subscriptionB. Healthy refreshing contentC. Public goodD. Financial inclusion

查看答案 / Show answer

答案 / Answer: A.

20. ChatGPT Plus 的 trilemma 是：What is the ChatGPT Plus trilemma?

A. IP/brand/contentB. Fixed price, unlimited usage, latest modelC. Logistics/payment/identityD. KOL/KOC/MCN

查看答案 / Show answer

答案 / Answer: B.

21. Xiaomi TV 的“被订阅”模式说明用户可能是：Xiaomi TV's “being subscribed” model suggests that the user may be:

A. The product sold to advertisersB. The bank regulatorC. The IP ownerD. The warehouse asset

查看答案 / Show answer

答案 / Answer: A.

22. JD Finance 与 GLP 的主要差异是：What is the main difference between JD Finance and GLP?

A. JD 从数据到信用资产，GLP 从实体资产到金融工具 / JD turns data into credit assets; GLP turns physical assets into financial instrumentsB. 两者都只卖广告 / Both only sell adsC. GLP 是创作者平台 / GLP is a creator platformD. JD 是 IP 公司 / JD is an IP company

查看答案 / Show answer

答案 / Answer: A.

23. Transient advantage 要求企业：Transient advantage requires firms to:

A. 永远守住一个产品 / Defend one product foreverB. 持续创造、利用、退出和重配优势 / Continuously create, exploit, exit, and reallocate advantagesC. 不做实验 / Avoid experimentationD. 只降低价格 / Only lower prices

查看答案 / Show answer

答案 / Answer: B.

24. “平台决定创作者类型”最贴近哪个说法？Which phrase best matches “the platform determines the type of creator”?

A. Platform determinismB. First sale doctrineC. Capital arbitrageD. Anti-commons

查看答案 / Show answer

答案 / Answer: A.

简答题预测与答题要点

每题下面先给中文答题思路，再给英文表达。考试时可以直接套英文句式。Each question gives the Chinese logic first and then English wording. You can reuse the English sentence patterns in the quiz.

1. Explain why platforms often beat pipelines in digital business.

答题点：平台连接多边用户，管道线性创造价值。平台拥有网络效应、零边际成本、数据反馈和生态创新。案例可用 Apple vs Nokia 或 Meituan。Answer: A platform connects multiple sides, while a pipeline creates value linearly. Platforms win through network effects, zero marginal cost, data feedback loops, and ecosystem innovation. The key advantage is orchestration and participation, not just product quality.

2. Why are operational moats harder to copy than technology moats?

答题点：技术功能可被模仿，但 JD 物流、美团骑手密度、商家 SaaS、履约调度需要长期资本、密度、流程和行为习惯。Answer: Technology features can be copied, but operational moats require long-term capital, density, processes, and habits. Use the four-moat test: fulfillment density, merchant lock-in, user habit durability, and subsidy sustainability.

3. Analyze Netflix's self-disruption capability.

答题点：Netflix 多次主动自我颠覆：DVD、流媒体、原创内容、广告/付费共享、直播/游戏生态。Blockbuster 因为 late fees 是利润中心，无法自我破坏。Answer: Netflix repeatedly cannibalized itself: DVD → streaming → originals → ads/paid sharing → live/games ecosystem. Its deepest moat is the organizational willingness to pivot before the crisis fully arrives.

4. Compare Disney and Netflix as business models.

答题点：Netflix 是平台/分发/注意力生意，主要靠订阅和广告，内容会折旧。Disney 是 IP 帝国，靠电影、流媒体、乐园、商品和授权，IP 可跨代际增值。Answer: Netflix is a platform/distribution/attention business based on subscriptions and ads; its content depreciates. Disney is an IP empire using film, streaming, parks, merchandise, and licensing; its IP appreciates across generations.

5. What makes IP last for 50+ years?

答题点：长寿 IP 需要 timeless emotion、controlled evolution、多代际入口、稀缺/不过度曝光、所有权和质量控制。Answer: Long-lasting IP requires timeless emotion, controlled evolution, multi-generational entry points, scarcity or anti-overexposure, and ownership plus quality control. Mickey, Hello Kitty, and Pokémon are proven examples; Pop Mart's durability is still unproven.

6. Why do different platforms produce different creator types?

答题点：YouTube 奖励长观看和广告观看，TikTok 奖励病毒传播和电商转化，小红书奖励购买意图和信任。Answer: YouTube rewards long watch time and ad views; TikTok rewards viral hooks, conversion, and Shop; Xiaohongshu rewards purchase intent and trust. Platform economics shapes creator destiny.

7. Can Li Jiaqi's live-commerce model be copied in the US?

答题点：需要 platform economics、culture、infrastructure 三层匹配。中国有移动支付、物流、平台内购物闭环、信任缺口和价格敏感直播文化；美国不一定具备。Answer: The model requires alignment among platform economics, culture, and infrastructure. China has mobile payment, logistics, embedded e-commerce loops, trust gaps, and price-sensitive live-sales culture. The US lacks some of these conditions, so the model cannot be copied directly.

8. How does FinTech solve the cost-risk-scale triangle?

答题点：传统金融服务小额贷款成本高、风险难判断、线下网点难扩张。FinTech 用移动开户、远程 KYC、替代数据、自动风控和平台分发降低成本。Answer: Traditional finance struggles with small-ticket loans because unit costs are high, risk is hard to assess, and physical branches do not scale. FinTech uses mobile onboarding, remote KYC, alternative data, automated underwriting, and platform distribution to reduce cost and improve risk assessment.

9. Why did regulators worry about Ant Group?

答题点：创新端是普惠金融、降低 SME 成本、高效信贷；稳定端是影子银行、资本套利、过度借贷、数据集中和系统性风险。Answer: The innovation argument is financial inclusion, lower SME lending costs, and efficient credit allocation. The stability concern is shadow banking, capital arbitrage, consumer over-borrowing, data concentration, and systemic risk.

10. Explain Waterdrop Mutual Aid's rise and closure.

答题点：低门槛、社交传播、互助共担帮助快速获客，但 mutual aid 主要是保险商城的 funnel。关闭原因是逆向选择、分摊成本上升、缺少准备金/再保险和监管不确定。Answer: Low entry cost, social sharing, and mutual risk pooling helped Waterdrop acquire users quickly, but mutual aid was mainly a funnel for the insurance marketplace. It failed because of adverse selection, rising per-share cost, lack of reserves/reinsurance, and regulatory uncertainty.

11. What is knowledge extractivism, and why does standard price theory struggle with it?

答题点：企业一次性购买/抓取知识后用 AI 永久复制。传统边际定价假设劳动和知识可以重复出售，但 AI 复制会消灭未来市场。Answer: Knowledge extractivism means a firm acquires knowledge once and uses AI to replicate it permanently. Standard marginal pricing assumes knowledge or labor can be sold repeatedly, but AI replication may destroy the future market for the seller.

12. Which solution to AI knowledge extraction is most feasible?

答题点：可选 data provenance labels、experience royalties 或 data trusts。关键是评价可行性和局限，不要说单一方案万能。Answer: Possible solutions include data provenance labels, experience royalties, or data trusts. A strong answer evaluates feasibility and limitations rather than claiming one solution can solve everything.

13. How do you judge whether a subscription is real or pseudo?

答题点：用 Subscription Triangle：停付损失什么？边际成本递减、平坦还是递增？关系是共创、服务、绑架还是攫取？Answer: Use the Subscription Triangle: what does the user lose if they stop paying? Is marginal cost decreasing, flat, or increasing? Is the relationship co-creative, servicing, hostage, or extractive?

14. Why is ChatGPT Plus a paradoxical subscription?

答题点：模型升级带来持续新价值，但重度使用会增加 GPU/inference 成本。固定价格、无限使用、最新模型三者不能同时成立。Answer: ChatGPT Plus offers refreshing value because models improve, but heavy usage increases marginal cost through GPU and inference expenses. Fixed price, unlimited use, and latest model access cannot all hold at once.

15. If asked to design a digital business strategy, what should you include?

答题点：写清 platform sides、control points、network effects、monetization model、data feedback loop、moat、监管/伦理风险和 transient advantage 路径。Answer: Identify platform sides, control points, network effects, monetization model, data feedback loop, moat, regulatory/ethical risks, and the path for transient advantage. Course cases make the answer stronger.

最后 5 分钟背诵版

平台胜出靠网络效应、零边际成本、数据反馈和生态控制点。Platforms win through network effects, zero marginal cost, data feedback, and ecosystem control points.

护城河不只是技术；运营密度、用户习惯、商家锁定和转换成本更难复制。A moat is not only technology; operational density, user habits, merchant lock-in, and switching costs are harder to copy.

内容会折旧，IP 会增值；创作者收入取决于平台经济、文化和基础设施。Content depreciates; IP appreciates. Creator income depends on platform economics, culture, and infrastructure.

FinTech 把支付信任和替代数据转化为信贷、保险、理财，但也带来监管和系统性风险。FinTech turns payment trust and alternative data into credit, insurance, and wealth management, but creates regulatory and systemic risks.

AI 知识攫取是一次性转移、永久复制、公司捕获剩余价值；可用原始积累、数字租、反公地解释。AI knowledge extraction involves one-time transfer, permanent replication, and surplus captured by the firm. Use primitive accumulation, digital rent, and anti-commons.

判断订阅：停付损失什么？边际成本如何变化？用户是客户还是产品？Subscription test: what is lost when stopping, what happens to marginal cost, and is the user the customer or the product?