The Lean Startup

This book was recommended to me by VC Michael Gibson after I asked him how I could learn about starting businesses. Jon Paris (from my small group) also vouched for the book, so I became really interested in reading it. The book is mostly about how to run a successful startup and prevent failure by following a quick, effective process. The five basic principles are:

1. Entrepreneurs are everywhere
2. Entrepreneurship is management
3. Use validated learning
4. Use innovation accounting
5. Use Build-Measure-Learn feedback loop

Vision

Many startups fail, but success can be engineered when following the right process. What Eric Ries proposes generally goes against what most of think about companies. For example, he defines entrepreneurship as a form of management, despite their opposite connotations (entrepreneurs hate management!).

Start

Use learning milestones to track progress rather than focusing on metrics that, while seemingly indicate progress, are not a sustainable measure to evaluate how the company is really doing.

Startups should focus on building the right thing (what customers will want and pay for) as quickly as possible. Instead of making a complex plan and following it for months, use the Build-Measure-Learn feedback loop to get results faster. As the name suggests, you want to quickly create/modify a product, run an experiment to see how customers react, and use the findings to make better improvements.

Your vision is the fundamental change you want to see in the world. Your strategy is how you will accomplish this goal, and includes the business model, product road map, assumptions on who customers are, and more. Your product is the result of the strategy and will constantly evolve as you optimize.

Define

Startup: a human institution designed to create a new product or service under conditions of extreme uncertainty

Innovation: though unpredictable, it can still be managed

Leadership: creating conditions that enable employees to do the kinds of experimentation that entrepreneurship requires

Learn

Customers do not know what they want in a product; asking them is therefore useless. Instead, figure out what they want by their actions or inactions (study how they interact with the product, usually by getting volunteers and conducting in-person tests). Any action that does not contribute to learning about what creates value to customers is a waste.

It's possible to build most products. The more important question, however, is "Should the product be built?" and "Can we build a sustainable business around this set of products and services?"

Experiment

Do not "Just do it" when it comes to launching a product. If you fail, you won't get the validated learning necessary to adapt your product. Instead, test a small idea to answer big questions (Zappos wanted to test the assumption that there was a demand for buying shoes online, so they took pictures of another store's shoes and sold them on their website).

Find early adopters, which are customers who most eagerly want the product. They tend to be more forgiving of mistakes and will be willing to give feedback.

Success is about learning to solve customer's problems, not just delivering new features.

An experiment is more than just a theoretical inquiry; it is also a first product. If any experiment is successful, managers can start enlisting early adopters and hiring employees.

Important Assumptions Every Startup Makes

1. Value Hypothesis: Does the product actually bring value to customers? Use experiments, not surveys, to prove it.
2. Growth Hypothesis: How will new customers discover the product?

Questions For Product Development

1. Do consumers recognize that they have the problem you are trying to solve?
2. If there was a solution, would they buy it?
3. Would they buy it from us?
4. Can we build a solution for that problem?

Steer

The products that a startup builds are actually experiments; learning how to build a sustainable business is the outcome. Use an MVP to experiment and test your most important assumptions (value and growth hypothesis). The MVP is the version of the product that allows the quickest cycle of the Build-Measure-Learn feedback loop while also expending the least effort. Even if you are on time and on budget, it doesn't matter if you build something no one wants.

For example, Facebook validated their value hypothesis after realizing that half of their users used the site everyday. It is clear that the product is valuable to customers.

Leap

Use a strategy to figure out the right questions to ask. A startup's early efforts should be focused on testing the right assumptions as quickly as possible.

Successful entrepreneurs discover which parts of their plans are working and which aren't. Being "at the right time and place" is not enough.

Test

First products are not meant to be perfect. In fact, early adopters can be suspicious when a product is too polished. Therefore, adding additional features beyond what is needed for the experiments is wasteful, especially when it slows down the experiments. Depending on what industry you're in, the MVP can be super complicated (like a working prototype) or as simple as an ad. Focus on what assumptions you need to test and how you'll test them.

If you don't know who the customer is, you don't know what quality is. Ask them what features in the design they care about, as it may differ from what you think matters most.

The only way to win is to learn faster than anyone else. Remove features from the MVP that do not contribute directly to the learning you seek, and start the process of validated learning as soon as possible.

Types of MVPs

• Video MVP: Explain the concept and ask people to sign up.
• Concierge MVP: Do everything yourself to create an awesome, personalized experience. For example, a grocery delivery app might have the CEO personally delivering it to a customer to make the experience the best for the customer.
• Wizard of Oz MVP: Simulate the technology you would need to develop by doing it yourself behind the scenes.

Measure

How do you know if you are making the product better? A good design is one that will change customer behavior for the better. To measure, you need a clear baseline metric, a hypothesis about what will improve that metric, and a set of experiments to test that hypothesis. Split testing is a good way to test different versions of a product, letting you know what matters to them and what does not.

Progress is measured by evidence that a sustainable business can be built around the product or service. Measure your productivity according to validated learning, not in terms of adding new features.

The 3 A's of Metrics

1. Actionable: Clear cause and effect between experiment and metric
2. Accessible: Make the reports as simple as possible so everyone understands them
3. Auditable: You must be able to test the data by hand

Innovation Accounting

Jon Paris told me that this is the most important part of the book. Innovation accounting is a way for startups to objectively prove they are learning how to grow a sustainable business. By getting rid of vanity metrics (misleading metrics that seem to suggest growth) and replacing them with learning milestones, can be kept accountable for their decisions.

First, establish an MVP to get real data on where the company is right now. Next, tune the engine "towards the ideal" by making all the optimizations and changes possible. When the possibilities are exhausted, the company must make a decision: pivot or preserve.

Pivot Or Preserve

The true measure of a runway is how many pivots a startup has left: the number of opportunities it has to make a fundamental change to its business strategy. When entrepreneurs have unclear hypotheses, it's almost impossible to experience complete failure, which is a necessary force to promote radical change and pivots.

Start now because you will never learn what needs to change without launching. Be careful though, since low-quality MVPs only work for early adopters; mainstream customers have higher demands for their products.

A pivot is a new strategic hypothesis that will require a new MVP to test. Optimization, tuning, and iteration are all methods to test a clear hypothesis in the service of the company's vision.

Types of Pivots

• Zoom-in: A single feature becomes the whole product.
• Zoom-out: What was considered the whole product becomes a feature of something much larger.
• Customer segment: We have built a product that solves a real problem for real customers, but they are not the customers we originally planned to serve.
• Customer need: The problem we were trying to solve turns out to not be very important. However, we discover related problems that are important in the process.
• Platform: Change from an application to a platform or vice versa
• Business architecture: High margin, low volume versus low margin, high volume. Or B2B versus B2C.
• Value capture: How you decide to make money.
• Engine of growth: Change in strategy to seek faster, more profitable growth.
• Channel: Change your distribution channel.
• Technology: Achieve the same solution with a different technology.

Accelerate

Batch

Testing in smaller batches leads to faster learning. The ability to learn faster from customers is an essential competitive advantage because it means you can roll out new features and validate them quicker. For example, I could quickly manufacture 100 products and learn from it even though I would get a much better price per for 1000 of them.

Grow

Your engine of growth is how startups achieve sustainable growth, which is when you get new customers coming in as the result of the actions of past customers. Measuring performance in the following engines of growth give good indicators on how well the startup is doing.

3 Engines of Growth

1. Sticky Engine: The rate of new customer acquisition exceeds the churn rate (percent of users who stopped using the product)
2. Viral Engine: The viral coefficient (how many more users join for each new member) is above 1
3. Paid Engine: The customer lifetime value outweigh the cost of advertisement

Adapt

Training your team to think "lean" is difficult, though important. Most mistakes are caused by flawed systems, not bad people. Always try to figure out how to prevent those situations in the future. To understand the root causes of failure, use the "Five Whys"

The Five Whys

1. Use it for every major problem
2. Setup a meeting with everyone involved, or who is affected by the problem
3. Ask "why did the problem happen?" and then ask why to each subsequent answer until you have reached the solution

Example:

Why did the website go down? The server crashed

Why did the server crash? The free tier expired on AWS

Why were we not notified about nearing the usage limits? We were, but the email was sent to an old company account.

Why was it sent to the old company account? We did not update our company information

Why did we not update our company information? Don't know. Maybe we should update the company information.

Innovate

Do not make the mistake of splitting off teams into parts like "frontend", "design", "communications". Oftentimes, your employees will need to be involved in multiple sections of the company to get their work done. Therefore, you want teams that are cross-functional, removing the need for your employees to go through ten different people to answer questions.

3 Structural Attributes For a Startup

1. Scarce but secure resources
2. Independent authority to develop the business
3. Personal stake in the company's outcome

Epilogue

Waste comes from working on the wrong things, not doing them inefficiently. Working on something useless at high speeds is completely wasteful, so don't do it.

Final Remarks (6/13/23)

I spit-balled a lot of vocab words, but what do they really mean? I really suggest picking up the book and reading it through; it will clarify a lot of the concepts that I summarized in this post.

This book was probably the most enjoyable and immediately applicable for me. I read this book about 3 months ago, before I had any idea what I would be doing with LunchTrak over the summer. Exposure to these ideas helps me clearly see what strategies I can be using to achieve validated learning, which is the most important idea I took away from the book (Jon Paris said innovation accounting and he is pretty successful). I like this idea mainly because it makes startups seem so much more simple: all they need to focus on is learning how customers react to their product and how they can get more products.

This was a pretty long post, but I'll update it in the future if I remember any important ideas that I missed.