《Lean Analytics》读后感1000字

话题：Lean Analytics 读后感 

　　《Lean Analytics》是一本由Alistair Croll / Benjamin Yoskov著作，O'Reilly Media出版的Hardcover图书，本书定价：USD 29.99，页数：440，特精心从网络上整理的一些读者的读后感，希望对大家能有帮助。

　　《Lean Analytics》精选点评：

　　●中间两章的案例蛮无聊的

　　●中间几张普及了基本概念的错误

　　●回答的一个核心问题是: 在不同业务模式下，应该看哪些核心指标。以及回答这个问题背后的方法论。 英文比较简单，读起来比较轻松。

　　●Lean系列里比较无聊的一本.

　　●非常棒的结构化产品数据分析入门书籍；根据不同领域提供的分析架构也非常sophisticated

　　●The introductory concept of MVP and extensive case sttudy.

　　●还算比较实用

　　●产品入门书籍，废话有些多，只有前几个章节比较有用。

　　●被书名骗了，不如叫精益商务分析吧，没时间就不用专门看了。虽然某些点还是有用的。

　　●这波 lean xxx 系列其实是在尝试把管理学的那套东西应用到互联网领域。抱着这种态度去再看书中里那些「模型」、「分析」、「数据」以及案例就挺自然了。这种应用某种程度上是适时的恰到好处，可惜却掩盖不了管理学科本身的乏味呀。呵呵。

　　《Lean Analytics》读后感(一)：怎样为产品做决策

　　适合读者：新手进阶（工作1个月以后读），还没工作的话就不必读了，会读不进去。

　　最开始觉得只是还算不错，后来读进去之后觉得很棒，主要是这个几个地方

　　- 涉及了非常多的metrics使用案例，很容易学习，尤其是对没有太多相关经验的团队而言.

　　- 对商业模式非常好的介绍，比如电商、UGC。一方面是我之前对这些不太懂，相当于系统性的梳理了一下，另一方面，每种模式的metrics关注点是有区别的，很实际。

　　- 最重要的，是案例中通过分析解决问题的思维方式，和依靠数据的决策方式。而不是随便想个idea，或者仅仅使用metrics。

　　如果时间不够，可以先看视频，貌似还算不错

　　《Lean Analytics》读后感(二)：创业初期如何靠数据分析来积累客户

　　Amazon评分还不错。评价摘录下面是

　　“This is one of these books that make you feel like you spent too little money for the value it provides. It's that good.

　　It is one of the most straightforward and comprehensive business books I've read. The information is clear, thorough and the case studies are not only interesting, but also very valuable since it allowed me to compare my business with other businesses. I now have a better feel of where we stands. The book stresses the importance of "actionable" metrics (vs vanity metrics) and helped me figure out the key metrics we should be tracking for our business.

　　Lean Analytics is an entertaining yet insightful read. Definitely worth every penny.”

　　《Lean Analytics》读后感(三)：操作手册里的星空视角——Lean Analytics "Minimum Viable Vision"小节阅读偶得

LEAN ANALYTICS这本书是多年之前买的，最近出于强化英语的目的在重看（汗），看着看着却有种“初闻不识曲中意，再听已是曲中人”的感觉。这本书的框架和内容其实很有前瞻性，但如果没有基础的产品或数据分析经历打底，大概也只是看个热闹。曾经我是个看热闹的，时至今日也不敢说有多深入的体会，但这轮重读，确实激发了一些从前没有的思考。

书中最被推崇，也是个人认为最有价值的，可以说是6大典型互联网产品的数据指标体系，包括SaaS，电商，媒体，移动应用，UGC内容平台和双边市场平台（也就是一度很火的共享经济吧）。当然这是大约7、8年前的视角，如今常见的产品体系已不限于此，也有很多产品是这6大类的综合体。不过，对每一类来说，关键的数据着力点、以及在业务发展的每一步需要关注的具体指标，书中都有很戳中膝盖的解读，有些地方甚至详细到手把手教。虽然时代有变，但万变不离其宗，这些认知点，无论什么时候回头来看，都有一定的启发性。

（所以还是推荐入了门的同学们好好读一读本书，有中文版，不用全看，和自己工作相关的商业模式、和自己工作相关的产品发展阶段那部分内容看明白就挺好。）

不过，这次阅读中，我个人印象最深的却不是这些了，而是一个讲“Minimum Viable Vision”的段落（英文版 P217 - 219）。这个概念对应的，是现在大家已耳熟能详的“Minimum Viable Product”（MVP-最小可用产品）。作者引用了一位采访对象的论述：如果想创建一个伟大的公司，比一步一步找到、优化MVP更重要的，是找到这个“MVV”。

要说“Vision”这个词，似乎有点遥远，又有点虚妄。尤其在眼下，这个产品、人、潮流的任何方面都早已“指标化”的时代，好像不找到一个肉眼可见、路径可落实的数据机制，就不够资格来讨论产品；而“视野”，无形而无用。的确，在技术成熟、商业模式精深发展的行业中，数据驱动的操作和决策是不可或缺的，也是最能带来直接收益的；同时，在巨型互联网企业中，无论是一个人，一个team还是一个大部门，只要能在局部最优的方向上前进一步，在竞争及向上、向下管理等重要问题上都能具备一些主动性，实在是划算的事情。

但在细节数据指标的方向上用力，久了会有困惑。如何确定某种比较是否有意义？最“有效”的究竟是不是最“好”的方案？不同板块甚至互相矛盾的数据应当如何串联，又如何协调统一？不能跳脱一些来理解更大的生态的话，疑惑会越来越多。而对于这些问题，Minimum Viable Vision所考虑的要点，似乎给出了某种解答。

书中观点指出，一个好的Minimum Viable Vision，包含如下内容：

· You’re building a platform.

· You have recurring ways to make money.

· You’ve got naturally tiered pricing.

· You’re tied to a disruptive change.

· Adopters automatically become advocates.

· You can create a bidding war.

· You’re riding an environmental change.

· You’ve got a sustainable unfair advantage.

· Your marginal costs trend to zero.

· There are inherent network effects in the model.

· You have several ways to monetize.

· You make money when your customers make money.

· An ecosystem will form around you.

——好的，这些一点也不Minimum。观点的阐述者是一位孵化器的合伙人，其实很明显可以看出，这些Vision，都是对于“能搭建新生态”级别的startup的期望，可以说非常的，殷切，又远大。

它们离实操的数据指标已经有点远了，甚至远得和本书的整体格调不大相符（Lean Analytics可是一本操作性极强、极“干”的书）；但不得不说，如果能结合到具体的业务和指标上，这里面的每一点，都可以指导对数据体系的理解，延伸对指标体系的设计和修正。也就是，引领从看见树木到看见森林的这个过程。

比如很多国外公司做工具，一直是照着naturally tiered pricing这个方向去的；SaaS和各种对内对外的数据服务，终极目标就是make money when your customers make money，过程中的功能优化和客户pitch大都以这个为指导；做增长的或许可以在争取advocates 和inherent network effects上下功夫，这和获客成本密切相关，同时也指向重要概念leading indicator……

通过Vision去“回访”日常工作中接触的指标和产品逻辑，也许会得到更通透的推进思路。

不过，最终Vision还是要落回Product和data，“能够衡量”始终是本书的核心要求。只是在抬头看天和低头走路的过程中，作为从业者，尊重生态和商业的本性，也许能行进得更明白一点。

阅读中的小小收获，共勉。

　　《Lean Analytics》读后感(四)：如果说lean startup讲的是方法论，那么这本就是方法

　　定位嘛，就是产品经理入门级读物了吧。不过也不用每章都读。

　　前面的一些方式方法的问题搞清楚，后面分开商业模型，每一类可以选取着读。

　　数据需要定性还是定量？Initially, you're looking for qualitative data. 在产品还没有庞大的稳定的用户群体时，定性分析是关键。

　　好数据应该是什么样子的呢？作者定出了一些vanity metrics, they are up-to-the-right and shed no light on anything. 好的数据首先是干净的，然后是可比的，以百分比的形式，衡量增长变化，其次要按照时间划分好，便于明晰某个时间点上，产品的的变化，所导致的用户行为上的变化。理想的情况下，随着产品设计的改变，数据上的统计方法也是一起设计好的，并且数据的结果应该有什么样的产品策略改变，也是清楚的。

　　但是，让团队内的每一个人，对于产品应该做的事情达成一致就已经很耗费时间了，更别说获取数据的干净程度，开发上能不能有无限的人力，都很难说。假设，干净的数据取来了，也并不意味着得到了因果关系。Usually causations are simple one-to-one relationships. Many factors conspire to cause something. ... You prove causality by finding a correlation, then running an experiment in which you control the other variables and measure the difference. 谁都知道因果关系好，可是现有的科学知识，都只是针对自然世界的因果关系推导。市场，用户，产品这三个都是变数没有特定规律，只能谈经验的东西。作者书里说，"unknown unknowns" are where the magic lives. They lead down plenty of wrong paths, and hopefully toward some kind of "eureka!" moment when the idea falls into place. 看来产品的魅力（也即，从业者们一切痛苦的根源）就在于这个"unknown unknowns"了。为了验证因果关系，测试的方法呢，有：segments, cohorts, A/ B testing, and multivariate analysis.

　　“You might think that people will play your multiplayer game, only to discover that they’re using you as a photo upload service. Unlikely？ That’s how Flickr got started.”

　　一切都要依靠数据么？“Humans do inspiration; machines do validation.” 用数据的方式可以帮我们找到局部最优，产品经理则需要找到全局最优。过度地依赖数据，很有可能会失去上帝视角。

　　lines in the sand？书中提到对于数据的统计结果，是要有预期的，并不能拿来数据看一看就做决策。怎么画这道线呢？

　　1）当40%的用户说，没有你他们会很失望，产品可以进入扩张期。

　　2）创业之前，至少找15个人，谈谈你的想法，再冲出门去找投资。

　　太多需求和问题，先做哪个后做哪个？作者用了一个squeeze toy来比喻，按照more stuff，more people，more often，more money，more efficient，从上往下挤。尤其在产品初期，将精力汇聚在一个目标上非常关键。

　　quot;UGC is all about turning visitors into creators." 想像一下明天tumblr flickr等等都不再有任何内容更新，大家仅仅根据已有的内容进行互动，就不难发现UGC的关键真的是内容创建，不是别的任何。怎么赚钱呢？“The UGC business might focus on user contribution above all else, but it still pays its bills with advertising most of the time.” “商业模式”是个有魔力的词汇，好像很多人都喜欢创造新的赚钱方法，只要UGC能通过广告赚到足够多的钱，何必为了创新而创新呢。书里面有一个business model flip book，供读者遍历每一种商业模式，看看怎么赚钱适合自己。

　　我要怎么用数据呢？对于创业公司，数据并不是在探索新机会，提示功能改进，而是检测自己的假设。“In a startup, your business model — and proof that your assumptions are reasonably accurate — is far more important than your business plan. ”看来TDD不仅仅只是程序管理方法，也适用于公司本身。“Deciding what business you’re in is usually quite easy. Deciding on the stage you’re at is complicated. This is where founders tend to lie to themselves. They believe they’re further along than they really are.”

　　可以参考的内容也不仅限于数字。用户访谈的方法书里面也有介绍。如果公司里面没有专门的用户研究部门，可以看看书里的用研技巧介绍。人们对话时会有认同别人的倾向，我们通过对方的穿着、谈吐，洞悉对方的身份，说话的时候会照顾这种“身份”，更何况直接表明来意的用户访谈。促使用户说出真相也有一些小tricks：1）问相反的问题，看她是不是会说出你想听的话。2）让他付出过多，看他什么时候放弃。

　　书里对产品工作的不同frameworks做了总结，着重介绍了lean framework，以及何时才能从一个phase移动到下一个phase：

　　所以说此书还是最适合当工具书。每个phase都有一些干货，实际操作中，同一时间一个产品只会有一个焦点（如果有多个的话，则与书里面提出的OMTM原则相悖，所以真的不用通读此书）。

　　具体论点不再赘述，只做摘抄：

　　Empathy:

　　如前

　　tickiness Phase:

　　Don’t drive new traffic until you know you can turn that extra attention into engagement.

　　Expect to go through many iterations of your MVP before it’s time to shift your focus to customer acquisition.

　　Virality Phase:

　　viral coefficient

　　Revenue Phase:

　　Most people’s first instinct when things aren’t going incredibly well is to build more features.

　　the likelihood that any one feature is going to suddenly solve your customers’ problems is very small.

　　You’re moving from proving you have the right product to proving you have a real business.

　　more stuff to more people for more money more often more efficiently

　　In the Revenue stage, you need to figure out which “more” increases your revenues per engaged customer the most:

　　cale Phase:

　　If the Revenue stage was about proving a business, the Scale stage is about proving a market.

　　orter observed that firms with a large market share (Apple, Costco, Amazon) were often profitable, but so were those with a small market share (the coffee shop). The problem was companies that were neither small nor large. He termed this the “hole in the middle” problem — the challenge facing firms that are too big to adopt a niche strategy efficiently, but too small to compete on cost or scale.

　　caling is good if it brings in incremental revenue, but you have to watch for a decrease in engagement, a gradual saturation of the initial market, or a rising cost of customer acquisition. Changes in churn, segmented by channels, show whether you’re growing your most important asset — your customers — or hemorrhaging attention as you scale.

　　最后一个部分，才是全书的重点：how good is “good”？ 同样，也分了不同的商业模式来讨论。

　　当然用数据分析产品，也离不开产品上每个人的思维模式，数据是客观的，但是解读它的人仍然是主观的。前面细细碎碎一大堆，最后结尾回到了一些经验式的忠告和建议。

　　: "Our Lean Analytics stages suggest an order to the metrics you should focus on. The stages won’t apply perfectly to everyone. We’ll probably get yelled at for being so preｓｃｒｉｐｔive — in fact, we already have, as we’ve tested the material for the book online and in events. That’s OK; we have thick skins." 噗嗤

　　回头看来，这本的确非常preｓｃｒｉｐｔive，

　　《Lean Analytics》读后感(五)：全书内容摘要

　　本书结构：1. 基本分析 2. 根据business model和发展阶段确定metric 3. set target来确定各个阶段什么才是正常 4. 综合运用需要提前了解的概念：

customer development: 区别于以往"build it and they will come“的waterfall式建立产品和公司的方法，而是在产品和公司发展的每一步都持续地收集信息和反馈来进行方向性决策。start up: a organization formed to search for a scalable and repeatable business model"Lean" is about eliminating waste and move quickly - good for organization of any sizeLean Startup:Build ->Measure->Learn concept(one of core concept of learn startup):

　　art 1: Stop Lying to Yourself - why you need data to succeed

企业家不应该仅仅听从于gut instinct: Guts matter, but you need to test them with data因为很多数码资源和服务免费，所以初创企业的试错成本非常低，因而lean startup的模式开始流行Lean Analytics希望能戳破企业家对现实的扭曲认知，从而及时发现企业存在的客观问题，进而扭转颓势MVP vs. Concierge MVP: MVP is defined as Minimium Viable Product, the smallest thing you can build that will create the value you've promised to your market. Concierge MVP: instead of building MVP, manually run things behind the scene for the 1st few customers to test if the idea work - it's cheaper and faster than MVP. E.g. hand connecting driver and rider.Lesson learned from Airbnb concierge MVP: Sometimes growth comes from an unexpected aspect of your business. When you think up a worthwhile idea, decide how to test quickly with minimal investment. Define success metric beforehand and know what to do if test result proves right.Lean, analytical thinking is about asking the right question, and focusing on the one key metric that will produce the change you'are after.Segment: a group that share the same characteristic.Cohort: customer who joined at the same time. Cohort analysis help to see customer behavior in their lifecycle, can be done for revenue, churn, virality, support cost, etc.AB test/Multivariate test(test many factors at once)when choosing your business, find something you are good at, you want to do, and can make money from.Don't become a slave to data - big innovation needs human inspiration instead of machine optimization. Use data to test hypothesis, but don't rely on data to give you innovation.

　　What'a a good metric？ A number that will drive the changes you're looking for. A good metric is comparative, understandable, a ratio/rate, actionable.

Qualitative vs. Quantitative. Quantitative is better metric, but the beginning and ending phase usually involve qualitative data.Vanity vs. Actionable. You want your data to inform, guide and improve your business model and help decide actions. Ask yourself: "will I do differently based on this metric？"Exploratory vs. Reporting. Reporting data used to measure "known unknowns". Exploratory data used to explore "unknown unknowns". Explore is necessary when early-stage startup wants to find a magic areas that they can disrupt - e.g. find a segment of power users.Lagging vs. Leading. Lagging metric helps to find problem(though sometimes too late); leading metric helps to size opportunity, usually requires cohort analysis. Depending on the business problem, lagging metric for one problem might be leading for another.Correlated vs. Causal. Correlation helps to predict what will happen, but causation give opportunity to act on and change something. To prove causation, first find correlation, then run experiment.Draw lines in the sand(moving target): Adjust your metric as you learn more about your customer/business/industry

　　art 2: Finding the right metric for right now. Lean Analytics Framework5个阶段：

Empathy：MVP（Minimium Viable Product，最小可用产品）目标是找到一个用户需求，并且确保产品能满足用户需求。这个阶段早期主要通过用户调研来进行定性分析，后期才可能用到数据分析Stickiness：目标是改进产品，进而使用户愿意黏住. 根据retention/engagement来确定用户是否喜欢这个产品，进而不断改进。会用到cohort analysisVirality: 目标是运用有机的口碑营销和病毒式传播以实现增长。需要根据virality的3种类型来segment user。referral需要关注的是 病毒系数 和 病毒周期，如果一个用户会推荐两个用户成功注册，那么这个病毒系数就是 2，如果这是经过 1 年才做到的，那么病毒周期就是 0.5 年了。如果你的产品不能通过病毒传播，可以看NPS（Net Prompter Score，净推荐值）。假设我们有 100 个客户，这 100 个客户中，有多少是给我们推荐了新客户的，有多少只是自己用的，有多少是通过私下或公开途径对我们进行负面评论的？拿推荐我们的减去否定我们的，就是 NPS。好产品NPS应该在50以上Revenue: 目标是盈利。metric shifts from usage patterns to business ratios.关注的重点成了 profit on acquisition cost, LTV（Life Time Value）、CAC（Customer Acquisition Cost）、渠道分成比例、渠道用户盈利周期、成本等。警惕花钱买增长等不可持续发展的模式，同时使用上面三个步骤寻找新的增长点。Scale: 利用规模效应做大做强，确定自己想efficiency-focused降低cost还是differentiation-focused增加revenue；同时放眼整个行业，建立ecosystem；放眼全球，寻找新的市场；寻找发展机会。key metric包括compensation, API traffic, channel relationship, competitors. 因为规模太大所以可能需要不止一个metric，可以建立a hierarchy of metric.

　　创业公司在不同发展阶段有不同的KPI，但是每个阶段必须确定一个最主要的KPI, 因为

it answers the mos important question you haveit forces you to draw a line in the sandit focuses the entire companyit inspires a culture of exprimentation

　　usiness growth comes from selling more stuff to more people more often for more money more efficiently. Not all customers are good. 需要定义出你真正想要的customer，不要浪费钱和时间来获取gamer或者metric不达标的customer.

　　usiness Model 1: E-commerce

　　首先要确定你是注重loyalty还是acquisition，这决定了你marketing和产品设计的很多策略

　　最重要的metric是revenue per customer，而不是conversion rate/repeat purchase/transaction size

　　不要忽视货物管理，比如shipping, warehouse logistics, inventory

　　用户偏爱搜索而不是浏览，Users mainly use Search instead of navigation: keyword become more important

　　E-commerce: when visitor buy something from a web-based retailer. e.g. Amazon, Walmart.com, Expedia.

　　Customer funnel: arrive -> navigate to item -> click "buy" -> provide payment -> complete purchase, however, below are the new trends:

　　Recommendation engines are commonly used

　　Retailers segment traffic to optimize performance through test

　　urchases begin outside of the website: e.g. social network, email inboxes, online communities

　　Depending on retention rate, e-commerce can be categorized as acquisition mode, hybrid mode or loyalty mode. Loyalty-focused retailers like Amazon build recurrent relationship with users.

　　usiness Model 2: Software as a Service SaaS包含IaaS/Paas

　　主要原则是通过建立loyalty来避免churn，所以需要提前很久monitor engagement metric

　　aaS company offer software on an on-demand basis. e.g. Salesforce, Gmail, Basecamp, Asana, Skype, Dropbox

　　Revenues mostly from subｓｃｒｉｐｔion, some charge on consumption basis (e.g. storage/bandwidth).但是其他monetization方式也许有奇效

　　ricing models are tiered.

　　Many user freemium model of customer acquisition because of negligible viral acquisition cost, but you need to use carefully

　　usiness Model 3: Free Mobile App

　　ways to monetize: downloadable content(e.g. extra map for game play), in-character gaming content, advantages(e.g. better weapon), saving time, elimination of countdown timer, upselling to a paid version, In-game ads

　　大部分收入来源于小部分用户，所以主要原则是做好用户segmentation，主要metric是average revenue per user/paying user。和SaaS类似，也需要engage user，extract from subｓｃｒｉｐｔion, prevent churn.

　　usiness Model 4: Media Site

　　代表性的有Google search, CNET, CNN's website

　　主要收入来源是广告，形式包括display, pay per view, pay per click, affiliate models

　　核心竞争力是吸引user停留更久和浏览更多内容，需要注意保持广告和用户吸引力的平衡

　　usiness Model 5: User-Generated Content(UGC)

　　engagement最重要，主要原则是让更多看客变成内容生产者，并且让好的内容战胜坏的(e.g. fraud)，revenue会随之而来

　　公司包括Facebook, Twitter, Reddit, Wikipeida, Youtube

　　用户浏览和发布的比例是80/20。为了提高engagement，可能需要一些interruption的手段，比如email notification

　　需要注重以下metrics:

前期需要解决鸡生蛋蛋生鸡问题。吸引buyer更重要，因为buyer是消费者，有消费者之后更容易吸引厂家。可以seed - 自己创造供应(e.g. 3rd party news, drivers hired by Uber)成长期需要保证供应可以在各种维度上满足需求，e.g. 观察buyer和inventory的增长速率成熟期需要解决的问题包括，通过调控保证供需平衡，通过信用系统保证市场效率，避免用户私下交易，帮助seller定价

　　收入来源可以是每笔交易抽成，或者提供帮助卖家的服务可以用的metric:

marketplace efficiency -liquidity as “the percentage of listings that lead to transactions within a certain time period”,provider-to-customer ratio, and repeat purchase ratiosupply -Gross Merchandise Volume (GMV)user funnel- AAARRR metrics e.g. MAU,Time spent on site, funnel (bounce rate),Customer Acquisition Cost (CAC)

　　不同business model在不同阶段最重要的metric见书267页 Part 3: Lines in the Sand 所有business共通的几个指标：

Growth Rate - 5% per week: B2C的startup的growth rate一半会经历慢->快->慢三个阶段。不应该在确保产品MVP，sticky和viral之前就focus在growth上，因为这种growth往往是不可持续的花钱买增长。早期track user growth，中后期track revenue growth.# of engaged visitors - 30% MAU and 10% DAU:Pricing: you can use experiment to find the best price, beware of too high/low pricing or differentiated pricing causing discrimination concern. Need to understand the right tiers and elasticity of your product.一旦找到最佳price，可以降低10%来提高growth.Acquisition cost - <1/3 of customer's revenueVirality - >0.75: try to build inherent virality in the productMail list Effectiveness - 20-30% open rate and >5% CTRUptime and Reliability - 99.5% time up: be transparent to customers about the outageSite engagement - average 1 min per page(YMMV)Web performance - <5 second for page load

　　Metric Benchmark for E-Commerce Business Model:

Conversion Rate - 2% is normal, 10% is goodShopping cart abandonment - 65%Search Effectiveness: search effectiveness needs heavy invest because 9 out of 10 searchs lead to action

　　Metric Benchmark for SaaS Business Model:

Paid Enrollment - 2% try and 1% buy if ask for credit card upfront, 10% try and 2.5% buy if not ask credit card upfront: recommend not ask credit card upfrontFreemium vs. Paid - some prefer to lose 10% money for 20% more customerUpselling and growing revenue - 20% customer revenue growth each year, 2% paying customer to increase payment amount each yearChurn - <5% per month is must, 2% is good

　　Metric Benchmark for Free Mobile App Business Model:

Mobile Downloads - no typical benchmarkMobile Download Size - <50 MBCustomer Acquisition Cost - average <$0.75 per user, $0.5 for paid install and $2.5 for organic installApplication Launch Rate - significant amount of user never open your appActive User - significant percent of user never open for a 2nd time, but others use stablyPaying User - 2% paying user for Freemium model; 1.5% for in-app purchase modelRevenue per DAU - >$0.05Revenue per Paying User - $20 for Whales(10% of all players), $5 for Dolphins(40%), $1 for Minnows(50%)App Review/Rating Rate - 1.5% for paid app and 1% for free appFrequency of use and retention - vary by industry: monetization strategy可以根据这两个维度来制定，具体见书320页

　　Metric Benchmark for Media Site Business Model:

CTR - 0.5%-2%Session to Click Ratio - expect 5% clicks never appear on your siteEngaged time - >90 seconds on content page

　　Metric Benchmark for UGC Business Model:

Success upload rate - 42% for FB photo success uploadTime spent per day - 17 minEngagement funnel - 25% lurk, 60-70% do easy engagement, 5-15% create content

　　如果没有baseline，努力提升metric直到提升变得困难，这能帮你了解你的极限 Part 4: Putting Lean Analytics to Work

2b需要注意的metric:Ease of Customer Engagement and FeedbackPipeline for Initial Releases, Betas, and Proof-of-Concept TrialsStickiness and UsabilityIntegration CostsUser EngagementDisentanglementSupport CostsUser Groups and FeedbackPitch SuccessBarriers to Exit有几个特性需要注意：What you make may cannibalize the existing business, or threaten employees’ jobs.Inertia is real.大公司缓慢的惯性会成为一个阻力If you do your job well, you’ll disrupt the ecosystem.你的项目的生死的决定权在高层手里，有时他们会为了公司层面的考虑而做出对你项目不利的决策有时你没有disrupt industry而是通过创新让公司适应新的竞争和监管环境有时运用empathy来重新思考最基本的问题，能帮你把cash cow产品变成high growth industry，区别见BCG box:

　　How to install a Culture of Data in Your CompanyStart small, pick one thing and show valuemake sure goals are clearly understoodGet executives buy-inMake things simple to digestEnsure transparencyDon't eliminate your gutAsk good questions