(notes on) Lean Startup by Eric Ries
NB. These notes contribute towards my diyMBA, my attempt to learn lots without a formal MBA programme. Check out my full reading list, and why I’m doing it DIY.
Essential reading if you’re at the very start of your career, but those with more experience reading an overview is probably enough as the topics covered are so fundamental to how we all work in 2020, a testament to how revered this book is. For further reading check links at the end of these notes.
TLDR
“The only way to win is to learn faster than anyone else.”
Startups are different to established companies as they are often working upon unproven hypothesis. Those that can learn the fastest directly from users will win. To do this, test your assumptions early with the minimum investment possible (an MVP) with validated learning, and pivot if necessary.
My Notes
Why should a startup work differently to an established company?
Established management strategy relies on creating a plan, setting milestones, and delegating tasks. But this approach does not work for startups or anyone operating in new, unpredictable markets. They work like they’re preparing a space ship for take off, tinkering for years before the perfect takeoff.
Instead they must be inherently flexible, like driving a jeep across uncertain terrain, with a set goal in mind but able to respond quickly to obstacles.
First, find a sustainable business model through validated learning.
Learn fast how to acquire customers and earn money from them. Without achieving this first, your business will fail. You’ll achieve this through constant validated learning, not through planning. Set hypothesis and test them with customers, ideally by offering them some form of your product.
Value Hypothesis — does a product deliver value to its first customers?
Growth Hypothesis — will it find a bigger market later?
Use an MVP to test in market
The Minimum Viable Product is simple, offering a realistic experience of how your product works to enable early feedback. It can even be a fake product. Build, Measure, Learn as fast as possible to find your sustainable business model.
eg Drew Houston did not build Dropbox to test his hypothesis, but instead created a video of how it was meant to work. Waiting Lists exploded
Learn what works, and pivot if needed.
Split-tests show which products are valuable and which are wasted efforts, by withholding specific products from some users and comparing usage. If your products are not working, pivot away from your core hypothesis.
Focus on one growth engine
Sticky Engine — retain existing customers & increase engagement
Viral Engine — existing customers market your product
Paid Engine — paid marketing, CPA must be lower than ARPU
Good vs Bad Metrics
Find the best core metrics for you, using cohorts to compare different groups (eg new vs old users).
Vanity Metrics — social media likes, press view, hours spent working on a product — are misleading.
Further Reading
Design Thinking vs Lean Startup is a good intro to Design Thinking, which delays the building of an MVP until after a design team have first validated the need for a product and understood the customer’s underlying problem.
What Your Innovation Process Should Look Like by Steve Blank and Pete Newell, interesting take on how to adapt Lean Startup ideas and apply to larger corporations or government.
The Limits of the Lean Startup Method by Ted Ladd advocates for the strength of a solid strategy above a high volume of user tests, something missing from the Lean Startup methodology. These critiques are answered well in this piece by Andy Cars.
