Running Lean. Retelling one of the best startup books
Below is a free retelling of Running Lean by Ash Maurya. Amazon's rating is 4.8 out of 5. As far as I know, the book has not yet been translated into Russian.
Here is the essence of the book in a volume of less than 1% of the volume of the book, in my estimation. All questions are covered as I understood them, which does not necessarily coincide with the way the author wanted to convey them.
I admit that I’m violating someone’s rights: retelling the license does not require, but the illustrations are taken from the book. I wrote to the author in an attempt to clarify this question, but received no answer.
Kigigi author blog: practicetrumpstheory.com .
- Introduction to the Running Lean Process
- Three critical methodologies
- Running Lean Process Description
- Make “Plan A”
- Determine who customers
- Create business models
- Rank business models
- Test and change plan
- Get ready to experiment
- Risk Categories and Four Phases of Test Plan
- Phase 1. Understand the Problem
- Phase 2. Find a solution
- Phase 3. Validate Qualitatively
- Phase 4. Verify Quantitatively
- Make “Plan A”
- Three stages of a startup
- The ratio of the three stages of a startup to the four phases of testing a plan
- Keep the right focus
- Get financing on time
- Scale correctly
Introduction to the Running Lean Process
Running Lean is a systematic process for building a profitable business plan; a process that takes you from Plan A to a plan that works before you run out of resources.
Three critical methodologies
Running Lean uses, among others, the three most important methodologies: Customer Development, Lean Startup, Bootstrapping.
Customer Development - a methodology for continuously receiving feedback from the consumer, parallel to the product development process. The methodology was created by Steve Blank and can be found in his latest book, The Startup Owner's Manual: The Step-by-Step Guide for Building a Great Company. Key idea: “Get out of the building.” (Steve Blank)
Lean Startup- a methodology for using short fast iterations to test hypotheses. Lean Startup is a synthesis of Customer Development, Agile Software Development and Lean (Toyota Production System) methodologies. The methodology was created by Eric Ries, a description of which can be found in his book The Lean Startup. (The network has a retelling in Russian by Arkady Moreinis).
Bootstrapping is a methodology for the development of a company on the same money earned.
The Running Lean process consists of the following steps.
- Drawing up "Plan A". “Plan A” will consist of business models, each of which will consist of hypotheses.
- Systematic testing in the four phases of the hypotheses of your plan and a parallel change in the plan. Your plan here will turn from “Plan A” into a plan that will work.
Running Lean Process Description
Make “Plan A”
Determine who customers
Start drafting Plan A by identifying your potential customers. (Hereinafter: “customer” == “consumer” == “customer”, “user” == “user”).
- the consumer should always be in the center of your attention;
- separate consumers and users (an example is a search engine: a consumer is an advertiser, a user is the one who is looking for);
- Highlight small consumer segments.
Create business models
When planning, use the business model in the Lean Canvas format described below, not the business plan. The business plan is static, rigid and in most cases does not stand up to the test of the market. Lean Canvas is a one-page presentation of a business model, an adaptation of Alexander Osterwalder’s Business Model Canvas. Adaptation is to give an entrepreneur a better point of view (rather than an investor, for example) and to better take into account the risks of most online startups.
For each consumer segment, create your own business model by filling it in the order below. Each part of the business model is your hypothesis that will be tested.
List the top three problems of your clients and rank these problems by importance.
List existing alternatives.
Highlight early adopters.
Highlight consumers, users and all their roles.
Unique Value Proposition (UVP)
UVP - the essence of your product in a few words that can be placed in the title of the landing page (landing page).
Remember the following.
- Be different. But let your difference be meaningful, why do UVP a direct answer to problem number 1.
- Reach out to early followers.
- Focus on the end result or, in other words, on why the consumer buys the product. (An example is a resume writing service. “Professional design templates for resumes” is just a feature of the product. “An attractive resume that stands out from others” is just an intermediate result. “Dream job” is the end result and it should be included in the UVP .)
- Good formula for compiling UVP according to Dane Maxwell: Final result + Time to receive it + Counterargument to the objection. Example: “You have hot fresh pizza in 30 minutes, we don’t have time - you will get it for free.” (Instant Clarity Headline = End Result Customer Wants + Specific Period of Time + Address the Objections. Example: “Hot fresh pizza delivered to your door in 30 minutes or it's free.”)
- Choose your words carefully and make them a “part” of the product. (Examples. "Performance" - BMW. "Design" - Audi. "Prestige" - Mercedes).
- Answer UVP questions: what is your product? Who are your customers? Why do your customers need your product?
Create at this stage a high-concept pitch, a phrase that would clearly convey the essence of the product, building on existing solutions. Examples: YouTube - “Flickr for video”, “Aliens” (movie) - “Jaws in space”, Dogster - “Friendster for dogs”. High-concept pitch is not used for UVP, but may be useful in other circumstances when you need to convey the essence of your decision.
Do not rush to describe the solution. Quite often, it turns out to be significantly changed after the first interviews with potential consumers.
Inbound vs Outbound
Consumer access channels can be divided into two groups.
- Inbound Consumers find you themselves. Examples:
- Electronic books
- White papers
- Outbound You find consumers. Examples:
- Cold calling
Do not spend resources on the channels of the second group (Outbound) before you tested the business model. (Interviews are an exception).
Direct sales vs automatic sales
First, spend more time on direct sales, in which you have a personal contact with the consumer. This is the best source of feedback. ("First sell manually, then automate.")
Set a price right away if you intend to make money on the product. The reasons are as follows.
- Price is part of the product: the consumer perceives the product differently depending on the price.
- Price determines the consumer segment that you are targeting, and, accordingly, all other parts of the business model.
- Consumer willingness to pay is a form of validating the hypotheses of your business model.
The Freemium model is more likely not a business model, but a marketing tactic. As a rule, it is better to use the “Free Trial” approach.
Consider the cost of that version of the product that will be mature enough so that you can set a price for it.
Use the so-called Dave McClure's Pirate Metrics They are so called because their first letters form an exclamation “AARRR”. There are five such metrics, each denoting the share of customers (conversion) who have passed to the next stage in interaction with the product. The sequence of these steps is called a conversion funnel.
- Attraction (Acquisition) - the proportion of customers (potential) attracted through marketing channels and who have found an interest in the product (not immediately leaving the site).
- Activation - the proportion of customers who have gained positive experience using the product.
- Retention - the percentage of customers reusing a product.
- Revenue (Revenue) - the proportion of customers who pay for the product.
- Recommendation (Referral) - the proportion of customers who recommend a product (share a link, send invitations).
Each stage can be divided into sub-stages and conversion can also be measured for them. For example, at the stage of Attraction, you can take everyone who came to the web page for 100%, and then measure the proportion of potential customers who closed the site on the UVP page, with a tour of the product, with a price, etc. ...
Determine which metrics you will use and what exactly in your product will be measured by each of the metrics.
Remember the following.
- You should be able to trace the connection of any of your actions (changes in the landing page, for example) with the results (number of registrations, for example).
- You must be able to reach people for numbers. (For example, you should be able to write by email to everyone who has registered, but did not pay).
Unique Advantage (Unfair Advantage)
“A unique advantage is that it cannot be easily copied or purchased.” (Jason Cohen, A Smart Bear blog).
- insider information;
- support of a unique expert;
- dream Team;
- personal authority;
- existing customers;
- high positions in search results.
Rank business models
Rank business models from potentially more successful to potentially less successful. In assessing the success, consider the overall risk of erroneous hypotheses: a model more successful than the risk is less . Otherwise, you can have your own ideas about success: ease of implementation, possible turnover, possible marginality, etc. can play a role.
Uncertainty is the presence of more than one scenario.
Risk - uncertainty in which there is at least one negative scenario.
Risk is assessed by assessment.
- and the likelihood of a negative scenario
- and the magnitude of losses in such a scenario.
Your “Plan A” will consist of several possible business models, each of which is a set of hypotheses. Each hypothesis carries a risk.
Ranking Risk Assessment
Use the following order of parts of the business model to rank in which the risk they contain is reduced.
- Problem. (The importance of the problem to the consumer.)
- Channels (Ease of access to the consumer.)
- Income and expenses. (Margin value.)
- Consumer segments. (Their size.)
- Decision. (Technical ability to implement the solution.)
That is, for example, the highest risk of error in determining the importance of the problem for the user: if you make a mistake here, then everything else will not matter.
Seek advice at the business model ranking stage.
Your advisers may be
- potential client,
- potential investor
- another entrepreneur with special expertise, or practical knowledge that is relevant to you.
Ask specific questions. Examples are as follows.
- “What do you consider the most risky part of the plan?”
- “Did you cope with the risks? How?"
- “How would you test the risks?”
- “Do you know someone who could give advice on these issues?”
Remember about the “Paradox of a Consultant”: “Hire consultants for good advice, but do not follow the advice, but use the wisdom that you learn from them.” ("The Advisor Paradox: Hire advisors for good advice but don't follow it, apply it."). (Venture Hacks).
Test and change plan
- take the potentially most successful business model of the plan,
- test its hypotheses (that is, test them for truth in experiments),
- modify erroneous hypotheses.
So your plan will constantly change from “Plan A” to one that will work.
Get ready to experiment
An experiment is a cycle in which you realize an idea, measure the result and learn a lesson from the data.
Remember the following.
- Take the least possible step, after which it is possible to learn a lesson. (!!!)
- Formulate fake hypotheses (those that can be refuted).
- Track the relationship between the outcome and your specific actions. (This also applies to organizing the measurement of metrics).
- Focus on one key goal (metric) that you must achieve.
Examples of hypotheses
Bad: “Fame as an expert attracts early followers.” Good: “A blog post will give 100 new posts.”
Examples of experiment organization
Bad: create a full version of the product for the experiment to test its relevance. Good: to create something for the experiment, other than the product itself, which will test its relevance.
Conduct experiments as a team.
Remember the following.
- Ideal team size: 2-3 people.
- Mandatory competencies of the team: marketing, development, design.
Use interviews as your primary experiment tool.
Remember the following.
- An interview with potential customers is a study of what you do not know, what you do not know.
- Therefore, the questionnaire survey is only relevant for quantitative verification.
- Therefore, in an interview, you only establish the context, and then listen most of the time.
- Do not ask what the client wants, evaluate what he does. (Examples. 1. The client says that there is a problem, find out how he solves it. Perhaps it doesn’t. What does it mean, the problem is not so acute. 2. The client says that he would buy, offer to pay a part of the payment now and guarantee a refund funds.)
- Придерживайтесь одного и того же сценария проведения интервью. (Чтобы отследить связь между результатом и конкретными действиями. По-другому проведенное интервью — другой результат).
- Говорите сначала с самыми разными людьми, не ограничивайтесь только ранними последователями.
- Рассчитывайте на 20-30 минут на одно интервью.
- Рассчитывайте, что придется опросить 30-60 человек. (Прекращайте, когда новые интервью не будут давать новой информации).
- Документируйте (шаблонно) результаты интервью сразу после его завершения.
- Благодарите за интервью (скидкой, ранним доступом к сервису и т.д..).
- Используйте те каналы доступа к интервьюируемым, которые Вы включили в бизнес-модель.
- Проводите интервью с помощником, это поможет быть объективным.
- Проводите интервью по модели AIDA: Attention, Interest, Desire, Action.
Share your experiment with all team members. This is best done through an always accessible, constantly updated dashboard. It is convenient to include in it:
- a business model with a highlighted portion of it being tested,
- hypotheses tested in the experiment (there may be several for one experiment),
- measurement results
- lessons learned
- further actions.
Risk Categories and Four Phases of Test Plan
It is advantageous to first test the part of the business model that has the highest risk. At the same time, it is profitable, by conducting another experiment to test parts with the highest risk, to test other parts that are possible and appropriate to test in this experiment. Based on the foregoing, firstly, it is convenient to distinguish three categories of risks:
- Product risks (is the right product being built) (Product Risk);
- Consumer risks (are there any ways to the client) (Customer Risk);
- Market risks (whether a working business is possible) (Market Risk).
Each part of the business model will contain a risk of a certain category. Secondly, it is convenient to divide testing into four phases, in each of which parts of the business model with risks from each category will be tested. So in four phases the entire business model will be tested. Graphically, it looks as follows.
The unique advantage is not affected in any of the testing phases, because it can only be tested in real competition.
So, test the plan in the four phases described below.
Phase 1. Understand the Problem
Understand the consumer's worldview before confidently formulating solutions to his problems.
- problem-oriented landing page (landing page), blog post, advertising in a search engine or social network (all this can help collect feedback);
- Design Thinking Methodology (Rapid Contextual Design by Karen Holtzblatt, Jessamyn Wendell, and Shelley Wood);
- Methodology “User Centric Design” (Human-Centered Design Toolkit by IDEO).
You can use all the tools, but the interview is required.
Interview on the issue
What you need to know
- Who are the consumers. (Part of the business model - Consumer segments). (Risk Group - Consumer Risks).
- What kind of problems do consumers have. How they rank them (top 3). (Part of the business model - Problem). (Risk Group - Product Risks).
- How consumers are solving the problem now. Who are the competitors. (Part of the business model - Problem). (Risk Group - Market Risks).
Outline of an approximate scenario of an interview on a problem
Phase 2. Find a solution
Test the solution in an interview.
Use in this experiment not a finished product, but a demo version (prototype design, for example).
Interview by decision
What you should know
- Who are the early followers. (Part of the business model - Consumer segments). (Risk Group - Consumer Risks).
- How to solve consumer problems. What is the minimum set of features (for early followers). (Part of the business model - Solution). (Risk Group - Product Risks).
- Will consumers pay how much and how much. (Part of the business model - Revenue). (Risk Group - Market Risks).
Outline of an approximate scenario of an interview for a decision
Phase 3. Validate Qualitatively
In a quality study, be sure to create a sought after product.
- consumer life cycle validation.
Use both tools.
For this phase you should have the following.
- MVP Minimum viable product ( the Minimum Viable the Product, MVP ) - this is the least of what you can build, but it already offers consumers some value and that the consumer will be ready to pay.
- Organized by Continuous Deployment. In creating a software product, this methodology allows you to implement changes as soon as possible.
- Organized metrics measurement.
- Selling site.
This interview should include usability testing. (For more on usability testing, see Rocket Surgery Made Easy by Steve Krug.)
What you need to know
- Can you reach enough consumers. (Part of the business model is Channels). (Risk Group - Consumer Risks).
- What should be UVP. Do consumers reach the activation stage. Does MVP justify your UVP. (Part of the business model is UVP). (Risk Group - Product Risks).
- Do consumers pay. Is monetization right? (Part of the business model - Revenue). (Risk Group - Market Risks).
Outline of an MVP Interview Scenario
Consumer Life Cycle Validation
Ensure that 80% of early followers complete all stages of the conversion funnel. You selected these consumers with special care at the stages of testing a business model, so this percentage is so high.
Achieving the goal of 80%, continue to receive feedback in person or use the telephone line to support users.
Phase 4. Verify Quantitatively
In a quantitative study on a wide audience, be sure to create a product in demand.
Prepare to make changes to the product before launching it for a wide audience, since after launching errors and requests for new functionality will spill over.
Remember the following.
- The introduction of new functionality should be justified. ("Features must be pulled, not pushed").
- 80% of the time should be spent on improving the existing functionality and 20% on a new one.
- The introduction of new functionality should be validated in experiments in the same way as MVP was validated.
- Use the dashboard from the Kanban Dashboard to monitor the implementation of new functionality.
Sean Ellis Test
Use the following simple question to determine if there is sufficient initial demand for a startup product. (Posted by Sean Ellis).
“What would you feel if you could no longer use the company's product?” Answer Options:
- Extreme disappointment.
- A little disappointment.
- I would not feel disappointed. (Not such a necessary product).
- I no longer use the company's product.
If more than 40% of respondents say that they would be extremely disappointed with the inability to use the product, then this is a sufficient sign that you can ensure the further growth of the company.
An indicator of 40% was obtained by interviewing consumers of hundreds of startups.
Achieve this indicator.
Consumer Retention Test
Apply the same 40% to the Conversion Funnel Retention metric: you have enough initial demand for further growth if you retain 40% of your "activated" users (here the Activation metric is taken as 100%).
The indicator does not apply to the Income metric because it can give a false positive result: there are users who will pay, but will not use the product (because the payment card is attached to the account and auto payment is set, for example).
Three stages of a startup
Startup goes through three stages in development.
- Finding a solution to a problem (Problem / Solution Fit) - You make sure that people really have a problem, and you have a solution.
- Search for a product for the market (Product / Market Fit) - You are convinced that your solution is a product that customers are willing to buy.
- Scale - Scale up your business.
The ratio of the three stages of a startup to the four phases of testing a plan
The first two phases of testing a business model are in the first stage of startup development. The third and fourth phases of testing a business model - at the second stage of startup development.
Keep the right focus
In the first two stages of startup development, focus on your own training, that is, on finding a plan that will work. Experiments at these two stages often lead to significant changes in the business model. This change is called Pivot (a term coined by Eric Rhys).
In the third stage, focus on startup growth. Experiments at this stage lead to the optimization of a business model aimed at accelerating growth.
Get financing on time
Attract external capital after going through the second stage of startup development. Passing the second stage is the first serious achievement of a startup. Here you have a scalable profitable business model, that is, a working plan for a profitable business. At this moment, both you and the investor have one goal (one focus) - growth.
In the third stage, select the main growth mechanism. Eric Rhys describes the following three.
- Sticky - the business is growing due to the fact that from the period during the period of retained customers (Retention metric) more than those leaving (those who used it were counted as Retention metric for one of the previous periods, but stopped using it). An example is a SaaS solution, a client began to use it and it became part of his life.
- Viral - the business is growing due to the fact that your customers are actively sharing information about your product: on average, one customer should tell at least one new customer about your product. An example is a social network, a viral mechanism is part of its structure.
- Paid - the business grows due to the fact that part of the income from existing customers you invest in attracting new customers (Attraction metric), that is, in advertising, hiring sellers, etc. ... Recommended condition for using such a mechanism (according to David Skok): LTV > 3 * COCA (LTV - customer lifetime value (how much the customer brings for the whole time that he is your customer), COCA - cost of customer acquisition (cost of attracting one customer)).
You can use all the mechanisms, but it is beneficial to focus on one.