Review of Eric Ries’ ‘mandatory reading for entrepreneurs’.
This book has been on my wish list for some time now, as it has been recommended from all over the place. Not just recommended, but without noticing I had attended multiple workshops teaching the core concepts without realizing it. I received it last Christmas together with Surely You’re Joking, Mr. Feynman and now got around to actually reading it in early July.
It’s a myth that a great product, a brilliant team, amazing technology and an amazing idea at the right time are required to become successful. It’s actually a process, which means that it can be learned and thought. Entrepreneurship is a kind of Management - but while general management is enormously efficient, it does not apply to startups. A different form of management is required.
‘Lean Startup’ management is built on Five Basic Principles:
- Entrepreneurs are Everywhere - it will work with large companies and small ones
- Entrepreneurship is a form of Management - it is required from every Company needing innovation
- Validated Learning - frequent testing of ideas, continuous experimentation
- Build-Measure-Learn - All successful startup processes should accelerate this loop
- Innovation Accounting - focus on the boring stuff: measure progress, milestones and priorities
Part One: Vision
Every startup has two feedback loops:
- growth engine: Product/Core business model, where everything needs to work with precision
- steering: Regularly adjusting the path and acceleration
A plan can be executed flawlessly and still fail if assumptions are flawed (‘death by winning’). The goal of a startup is to figure out a valid business model step by step, testing and validating assumptions. Startups operate under conditions of extreme uncertainty.
To achieve and sustain innovation, an appropriate environment with appropriate metrics and incentives is required - most companies provide neither.
Validated learning is applying the scientific method to figuring out business models. Not every ‘learning’ is learning, it needs to be usable knowledge for your cause. Check with how customers behave, not just what they say.
The core question of ‘Lean’: Which efforts are value-creating, and which are wasteful? Everything that helps bring value to a customer has value at a startup (and indirectly, the parts required to keep the value-creating ones operational and in good condition). Measuring productivity in a startup-context is easy: how much validated learning did you get for how much effort?
“Just doing it” is achieving failure, you will succeed at seeing if it works, but likely neither that it works, nor why. Prefer experiments to interviews, and always build the smallest experiment possible.
You have to ‘leaps of faith’-assumptions that need to be tested:
- Value hypothesis: customers want this enough to pay money for
- Growth hypothesis: we can reach enough customers to finance ourselves
Running experiments early and often allows for feedback and massive improvements on the product.
Part Two: Steer
The main goal of a startup is to minimize the turnaround time of the Build-Measure-Learn cycle. Test your biggest/riskiest assumptions first.
If you have validated both your growth and value hypothesis, it’s ‘only a matter of scaling’.
The factors of a right place and time are only second to foresight and experimentation - but they win every time.
Genchi Gembutsu: Go and see for yourself - don’t rely on reports for critical decisions.
The goal of an MVP is to start the learning process - early adopters don’t need a fully working solution. Prototypes work fine - only do what you need to learn.
A startup needs to know they are actually improving their product, that they are making progress. For that, meticulously track your key metrics. Turning the growth engine, validated learning design changes have to improve metrics.
Running daily experiments is perfect. Map and model customer flows. Do a cohort analysis on your sales funnel for your key metrics. Don’t exclude people from your ‘target demographic’ because they don’t like your product. Experiments are much more successful if they are aligned with what customers really want. Optimizing the wrong product does not produce results. Execution with discipline still leads to failure, if the plan is flawed.
Don’t use vanity metrics (gross customer numbers), only metrics that will help you to act on them (enabling replication). Your metrics should be actionable, accessible, and auditable.
While ‘The Lean Startup’ is a rigid scientific methodology, ‘pivot or persevere’-decisions still require vision, intuition and personal judgment. Innovation accounting leads to faster pivots. Failure is part of learning. After a pivot, execution speed should improve, since infrastructure doesn’t need to be built again - but also, because you learned something critical about your customers.
The most important metric of a startup is how many pivots it can still make before it dies. The more, the more likely it is to survive (this is ‘the startup runway’). Hold regular meetings about pivoting. A pivot is not just a change, it’s a change designed to test new fundamental hypothesis about the product, business model, and engine of growth.
Part Three: Accelerate
Most decisions startups face are not clear-cut. It is paramount to test what creates value in the end.
Batching can help increase efficiencies, but only in higher volumes. Turnaround times for testing are usually worse, so introduce it only later. Notice failures in production processes as fast as possible. The most effective is to release fast and frequently in small batches. Use automated tests and quality insurance when possible. Fast learning cycles are what will keep your company alive.
There are multiple ways to grow - depending on advertisements is the worst of them, prefer word of mouth or usage side effects when possible. Knowing your lifetime customer value can help you make better marketing decisions. You should focus on just one engine of growth, multiple ones have complex interactions and are hard to model.
Product-Market fit is when a widespread customer base resonates with the product. If you have to ask if are there yet, then you are not.
Over time, the roles of everyone in your startup will change, though it’s hard to notice it. Introducing more bureaucracy with increasing company size is not the answer and will only slow you down. Never split the difference in arguments.
Having a decent training process for new hires is essential for them to start being productive. The requirement for standardized processes and concept curriculum will help you and everyone in your team to keep a clear view as well. Assign a mentor to new members. More important is, however, to have a methodical approach to evolving processes.
Not just the training process, every process needs to be evaluated and adapted regularly - conditions change all the time. Speed alone is not the goal, you need to be directed toward your goal as well. Do not neglect quality, or you will pay in required time and resources.
The five why’s can help you to gradually evolve your processes. Ask ‘Why’ five times when something went wrong, to uncover multiple layers of failure. Afterwards, invest proportionally to the errors’ severity in fixing all uncovered problems.
Be tolerant of mistakes the first time around, but never allow the same mistake to be made a second time. Someone with sufficient authority needs to oversee the five why’s process, or it will devolve. Decide on simple but iron-clad rules.
Remember that organizations have muscle-memory when making larger changes - involve impacted teams and customers as much as possible. Explain clearly why the change is needed and thoroughly communicate expectations from that change.
Internal or external startup teams require three structural attributes to innovate successfully:
- scarce but secure resources (small margins of error, but don’t let it be too many available resources)
- independent authority to develop their business (for fast iterating)
- a personal stake in the outcome (does not need to be financial)
Innovation-based teams and parent companies need to be protected from each other (from politicking, sabotage and other interference as well as from success threatening the parent companies core business). Create innovation sandboxes, where innovation teams just need to report about their experiments within certain parameters.
Always use the same metrics when comparing experiment results. This builds literacy and familiarity with the relevant parameters.
Different stages in the innovation cycle require different types of management. Allow team members to continue with the product or stay in their current stage.
Relying on brilliant employees will always give you worse results than building processes and systems that will improve the quality of everything. Find systems that allow every employee to be brilliant - it will help especially those that are brilliant already.
Avoid pseudoscience: always ensure that you know how the proposed cause and effect correlate.
There are multiple concepts and ideas I took away from this book, each one having a separate paragraph below:
I took part in a rapid prototyping workshop earlier this year. So the idea of building the smallest prototype to learn something was not new … but that a startup does only things like that (in the beginning) was quite the revelation for me. To be honest, startups might be even more fun than I thought.
New Perspective what a Startup is
I gained a new perspective on startups: the only goal is to find a sustainable business model. That’s it. After that, it’s ‘basically’ a normal scaling company with a bit of an innovation engine still running.
Extreme Application of the Scientific Method to Business
While I was aware of the scientific method and - at least partially - it’s application in business, ‘The Lean Startup’ provides a framework taking this application to the extreme. It’s probably the reason why it is so successful, even. I’m glad I don’t have to figure it out on my own - standing on the shoulders of giants sure is convenient.
Innovation Accounting of Validated Learning was also new to me, and a concept I’ll certainly use in the future. I’m sure that there will be enough opportunities to do so.
Cohort analysis seems so obvious and simple in retrospect. I still only grasped the basics, but it sure is a powerful way to create independent customer sections to run experiments on, and having a robust way of measuring effects.
Continuous Integration Everywhere
Continuous integration/testing is part of the software world I come from, but what is described as part of Lean methodology drives that to its extremes, and not just for software either. And I believe that gives me a good understanding why it works so well.
The Five Whys
When applied correctly, the methodology of asking ‘Why’ five times can be quite transformational, I believe. Nothing ever goes wrong on just the level it appears to - there are always much more fundamental problems involved, and finding and fixing them is among the most energy intensive and effective things one can do. I learned a lot from reading about them, and I can’t wait to apply this process.
Systems over People
Focusing on systems instead of brilliant employees is quite brilliant. Establishing processes that amplify employee effectiveness are among the most effective things anyone can do. I believe the main issue is to figure them out and enforcing them. Then, discipline creates freedom.
There were books recommended in different sections of the book. While both the Innovators Dilemma and the Innovators Solution were recommended in multiple chapters, The Principles of Scientific Management were only mentioned and recommended near the end. Still, I’d really like to read them soon.
- The Innovators Dilemma/Solution by Clayton Christensen
- The Principles of Scientific Management by Frederick Winslow Taylor
I really liked how the book was structured, and that the explained model got linearly more complex and intricate. The explanations were clear and the examples good to bring their respective points across.
The only thing I didn’t like was that early on a lot of forward-looking references were made (mainly to chapter seven, iirc). But the chapter didn’t reference any of the prior situations, so it kind of left these references hanging - or at least that was my perception.
Overall the book is well written, and I learned a lot from reading the book. While it was no litarical masterpiece, newly acquired concepts was worth reading it. It’s a good reference, and one I’ll likely use often in the future.
To be honest, I’d recommend it to everyone who creates value for a customer or the company - which is, I hope, everyone. Everyone should be able to get something out of it, and improve their work based on it. It will be, however, especially useful for everyone dealing with newly created companies or highly innovative teams.
Entrepreneurship is a kind of Management
[…] it’s the boring stuff that matters the most.
Startup success can be engineered by following the right process, which means it can be learned, which means it can be thought.
A startup is a human institution designed to create a new product or service under conditions of extreme uncertainty.
Unfortunately, “learning” is the oldest excuse in the book for a failure of execution.
The effort that is not absolutely necessary for learning what customers want can be eliminated.
This is one of the most important lessons of the scientific method: if you cannot fail, you cannot learn.
planning is a tool that only works in the presence of a long and stable operating history.
Customers don’t care how much time something takes to build. They care only if it serves their needs.
If a competitor can outexecute a startup once the idea is known, the startup is doomed anyway. […] Sooner or later, a successful startup will face competition from fast followers. […] The only way to win is to learn faster than anyone else.
piercing the reality distortion field is quite uncomfortable: commit to iteration
A solid process lays the foundation for a healthy culture, one where ideas are evaluated by merit and not by job title.
successful pivots put us on a path toward growing a sustainable business
The ability to learn faster from customers is the essential competitive advantage, that startups must possess.
Sustainable growth […]: New customers come from the actions of past customers.
The Five Whys approach acts as a natural speed regulator. The more problems you have, the more you invest in solutions to these problems.
of course, any innovation system eventually will become the victim of its own success.
There is a reason all past management revolutions have been led by engineers: management is human systems engineering.
There is surely nothing quite so useless as doing with great efficiency what should not be done at all. - Peter Drucker