分享一下《Lean Analytics》笔记

之前看到地里不少人推荐《Lean Analytics》这部书，在空余时间读了读，整理了一下笔记，分享给没空读的童鞋们～

我现在在一家start up工作，书中提及的metrics确实都是老板很重视，日常工作中会用到的，个人觉得该书还是挺适合new graduates。

书中讨论metrics部分按照E-Commerce，SaaS，Free Mobile App， Media Site，Media Site，User-Generated Content，Two-Sided Marketplaces做了区分，我现在还只读了SaaS部分，因为和自己所在的公司类型吻合。如果有感兴趣其他部分的，我之后也可以读了进行补充：）

-------------------------------------------------------------------------------------------------------------------------正文 笔记-------------------------------------------------------------------------------------------------------------------------
Part I: Stop lying to yourself
        ► What makes a good metric?
                ○ Comparative
                ○ Understandable
                ○ ratio or a rate
                ○ change the way you behave: avoid false metrics and make it valuable
        ► Metrics often come in pars
                ○ Conversion rate (the % of people who buy something) is tied to time-to-purchase (how long it takes someone to buy something)
        ► Five relations of metrics
                ○ Qualitative vs. quantitative
                        § Quantitative answers "what" and "how much"
                        § Qualitative answers "why"
                        § Combination is the key.
                ○ Vanity vs. actionable
                        § The point is that you're doing something based on the data you collect.
                ○ Exploratory vs. reporting
                ○ Leading vs. lagging
                        § I.e. in a B2B company, quarterly new product bookings are a lagging metric of sales success. By contrast, new qualified leads are a leading indicator, because they let you predict sales success ahead of time.
                ○ Correlated vs. causal
                        § Prove causality: finding a correlation, then running an experiment in which you control the other variables and measure the difference.
                        § Correlation is good; causality is great.
        ► Things to avoid as a Data Scientist
                ○ Assuming the data is clean
                ○ Not normalizing
                ○ Excluding outliers
                ○ Including outliers
                        § You probably want to exclude them when building data products. Otherwise, in a recommendation system, it will have the same items everywhere.
        ► Appendix
                ○ P28: the circle of life for analytical startups   

Part II: Finding the right metric
        ► One metric that matters (OMTM)
                ○ Capture everything, but focus on what's important   
        ► Lines in the sand
                ○ You need to pick a number and set it as the target.
                ○ Two methods:
                        § From your business model, which may tell you what a metric has to be.
                        § Look at what's normal or ideal. Knowing an industry baseline means you know what's likely to happen, and you can compare yourself to it.
        ► Model for SaaS
                ○ Key metrics
                        § Attention
                        § Enrollment
                        § Stickiness
                        § Conversion
                        § Revenue per customer
                        § Customer acquisition cost
                        § Virality
                        § Upselling
                        § Uptime and reliability
                        § Churn
                                □ In SaaS, churn is everything.
                        § Lifetime value
                These metrics follow a natural, logical order. You need to know which aspects of your business are too risky and then work to improve the metric that represents that risk.
                ○ Measuring engagement
                        § The ultimate metric for engagement is daily use. How many of your users use your product in a daily basis?
                        § For a non-daily use app, establishing a minimum baseline of engagement takes longer, the time it takes to iterate through a cycle of learning is longer, and harder to demonstrate enough value to keep people from churning.
        ► Churn
                ○ Can be measured weekly, monthly, quarterly, etc., but you should pick a timespan for all your metrics and stick to it in order to make comparing them easier.
                ○

                ○

                ○Churn complication: to fix that the number of churns in a particular period is affected by the entire period, but the number of customers at the beginning of a period is a moment-in-time snapshot. Simple version can give misleading results in startups where growth is varied or unusually fast. The following formula spreads out the total number of customers across the period.
                ○

        ► What stage are you at?
                1. Empathy
                        i. You need to get inside of your target market's head and be sure you're solving a problem people care about in a way someone will pay for.
                2. Stickiness
                3. Virality
                        i. Once you've got a product or service that's sticky, it's time to use word of mouth. Virality is also a force multiplier for paid promotion, so you want to get it right before you start spending money on customer acquisition through inorganic methods like advertising.
                4. Revenue
                5. Scale
                        i. With revenue coming in, it's time to move from growing your business to grow your market.

        ► Stage one: empathy
                ○ In the empathy stage, your focus is on gathering qualitative feedback, primarily through problem and solution interviews.
                ○ Later, you might score the interviews to keep track of which needs and solutions were of the greatest interest, because this will tell you what features need to be in your minimum viable product (MVP).
                ○ Find a problem to fix:
                        § The problem is painful enough
                        § Enough people care
                        § They're already trying to solve it
                                □ The current solution, whatever it is, will be your biggest competitor at first, because it's the path of least resistance for people.
        ► Stage two: stickiness
                ○ The focus now is squarely on retention and engagement. Segment these metrics by cohort to see if your changes convince additional users to behave differently.
                ○ Your top priority is to build a core set of features that get used regularly and successfully, even by a small group of initial users. Without that, you don’t have a solid enough foundation for growth.
                ○ The goal is retention.
                        § The more engaged that people are with your product, the more likely they'll stay.
                        § If the answer of "do we believe that the feature we want to build will improve stickiness" is "yes", figure out how to test that belief and start building the feature.
                        § Seven questions to ask yourself before building a feature.
                ○ How to handle user feedback
                        § User feedback suffers from horrible sampling bias. They reach out when they're ecstatic or furious.
                        § Don’t talk to just anybody. "You should group feedback from similar personas. Figure out who your customers are and focus your research on a particular type of person."
                ○ Minimum viable vision (MVV)
                        § The MVV you need in order to get funding demands a convincing explanation of how you can become a dominant, disruptive player in your market.
                ○ Problem-solution canvas
                        § Never underestimate the power of getting everyone on the same page-literally.
                        § Context on the canvas
                                □ Metrics/ current status
. 1point 3 acres                                □ Last week's lesson learned
                                □ Top problems
        ► Stage three: virality
                ○ The three ways things spread
                        § Inherent virality
                                □ Share the product by usage purposes.
                        § Artificial virality
                                □ Providing a bonus for spread.
                        § Word-of-mouth virality
                ○ Metrics for the viral phase
                        1. First calculate the invitation rate, which is the number of invites sent divided by the number of users you have.
                        2. Then calculate the acceptance rate, which is the number of signups or enrollments divided by the number of invites.
                        3. Then multiply the two together.
                ○ Attacking the leading indicator
                        § i.e. early on twitter focused its energy on increasing feed views because it knew its revenue would be tied to advertising - and that advertising could happen only when a user looked at her twitter feed. The number of feed views was a leading indicator of revenue potential even before the company hit the revenue stage.
        ► Stage four: revenue
                ○ Business model: it's pretty simple, really: look at the return on investment of your marketing dollar.
. .и                                

                                □ Your quarterly recurring revenue for quarter x (QRR[x])
                                □ Your quarterly recurring revenue for the quarter before x (QRR[x-1])
                                □Your sales and marketing expense for the quarter before x (QExpSM[x-1])
                ○ Customer lifetime value > customer acquisition cost
                        § When it comes to turning revenues into additional customers, the most basic rule is simple: spend less money acquiring customers than you get from them.
                        § The CLV-CAC math also needs to reflect the fact that there's a delay between paying to acquire customers and those customers paying you back.
                ○ Market/ product fit
                        § Most people's first instinct when things aren't going incredibly well is to build more features. Hopefully, we've demonstrated that this isn't the right approach, because the likelihood that anyone feature is going to suddenly solve your customers' problem is very small.
                ○ The breakeven lines in the sand
                        § You want to be breakeven - meaning your revenues exceed your costs on a regular basis.
                ○ Time to customer breakeven
                        § $27 Customer lifetime value
                        § 11 Months from activation to departure
                        § $2.45 Average revenue per customer per month
                        § $14 Cost to acquire a customer
                        § 5.7 Months to customer breakeven
                        You need 5.7 months' burn to keep the company running.
        ► Stage five: scale
                ○ Set three threes to follow: three big assumptions, three actions to take, and three experiments to run.
        ► Model + stage drives the metric you track
                ○ The core idea: by knowing the kind of business you are, and the stage you're at, you can track and optimize the One Metric That Matters to your startup right now.
                ○ Business model
        ► Appendix
                ○ P162 - 165 how to conduct a good interview
                ○ P166 -168 how to avoid leading the witness (Quite interesting. Suggest read.)
                ○ P208 - 210 Seven questions to ask yourself before building a feature
                ○ P267 - 269 Business model

Part III: Lines in the Sand
        ► Cancellation rate
                ○ One of the most critical metrics to track and understand.
                ○ Essential for calculating metrics like customer lifetime value.
                ○ An early warning signal that something is going wrong or that a competing solution has emerged.
                ○ Bottom line: 5% churn a month before looking at other things to optimize.
        ► CAC - CLV
                ○ 1:1 CAC to CLV ratio means they're spending all the money they make acquiring new users.
        ► Growth rate
                ○ Three distinct growth phase:
                        § Slow: where the organization is searching for a product and market to tackle
                        § Fast: where it has figured out how to make and sell it at scale
                        § Slow again: becomes a big company and encounters internal constraints or market saturation.
        ► Pricing metrics
                ○ If you charge too little, you'll arouse suspicion from buyers, who may wonder if you're up to no good or you're a scam.
                ○ If you charge too much, you may slow down the much-needed viral growth or take too long to achieve network effects that improve your product's functionality.
                ○ Price is related to what your customers are prepared to pay.
        ► Cost of customer acquisition
                ○ Bottom line
                        § Your acquisition cost should be less than a third of the total value a customer brings you over her lifetime.
                        § The CLV you've calculated is probably wrong. Churn has the biggest impact on CLV, and unfortunately, churn is a lagging indicator. It takes a long time to find out whether you've underestimated churn or overestimated customer spend.
                        § The acquisition cost is probably wrong, too. New customers incur an up-front cost - onboarding, adding more infrastructure, etc.
                        § Between the time that you spend money to acquire someone and the time you recoup that investment, you're basically "lending" the customer money. The longer it takes you to recoup the money, the more you'll need.
                        § Limiting yourself to a customer acquisition cost (CAC) of only a third of your CLV will force you to verify your acquisition costs sooner, which will make you more honest - so you'll recognize a mistake before it's too late.
        ► SaaS: Lines in the Sand
                ○ Paid enrollment
                        § Totango - asking for a credit card during signup means 0.5% to 2% of visitors sign up for a trial, while not asking for a credit card means 5% to 10% of visitors will enroll.
                        § Impact of requiring a credit card to try a SaaS product
        ► Appendix
                ○P300 Impact of requiring a credit card to try a SaaS product  

-------------------------------------------------------------------------------------------------------------------------结束 感恩------------------------------------------------------------------------------------------------------------------------

最后求个大米，感恩<3                                                                                                                                                                 

谢谢楼主，看来是本好书

谢谢楼主分享，今晚慢慢看

牛逼的总结，思路清晰，谢谢楼主！

马一个， 谢谢整理

马克吐温，感谢楼主。

马克一下，感谢楼主整理