The Lean Startup by Eric Ries - Summary and Book Review
The definitive book on running a Silicon-Valley style 'lean' startup. Must read for entrepreneurs.
The Lean Startup Book Review
Plenty of big startups use Lean Startup methodology to grow their business rapidly, and this book is the bible of the whole movement. Inspired by Lean Manufacturing and his work advising startups, Eric Ries developed his methodology. It's effectiveness spoke for itself, and the idea spread quickly through the startup world. If you want to know how startups grow so fast, this is the book to read.
The writing is pretty dry, and the book is structured in the tired 'thesis, example, example, example, conclude' format that plagues these types of books. Still, it's worth reading for the concepts.
If you hang out on Hacker News you probably won't pick up much from this book, though it could be a good review if you're about to start a company. Recommended for:
- The manager at a company with year-long product cycles and declining sales
- The two Silicon Valley startup CEOs who haven't read this book yet
- The entrepreneur in training who might start a company one day
Buy The Lean Startup on Amazon or check it out at your library.
Super Quick Summary
The point of a startup is to get through the Build-Measure-Learn cycle as quickly as possible to figure out what your customers will pay for. Quit using vanity metrics.
Quick Book Notes
The following are rough notes I took while reading. These are mostly paraphrased or quoted directly from the book.
Most startups fail.
Startup success can be engineered, it can be learned, it can be taught.
The Lean Startup Method
- Entrepeneurs are everywhere - Huge companies can (and should) have entrepeneurs
- Entrepeneurship is management
- Validated Learning - Startups exist to learn how to build a sustainable business
- Build-Measure-Learn - Run through this loop as fast as possible
- Innovation accounting - how to measure progress and prioritize work
The goal of a startup is to figure out the thing customers will pay for as quickly as possible
Startup: a human institution designed to create a new product or service under conditions of extreme uncertainty
If you don't test new things often, you get an office full of politians, fighting for their idea to make it into the release.
A company's only path to long-term growth is to build an "innovation factory" [see Purple Cow]
Learn
Validated Learning: demonstrate that a team has discovered valuable truths about a startups business prospects.
Don't ask customers what they want. Have them try something and measure their behavior.
Not, "can this product be built?"; better question is "should this product be built?"
Everything a startup does is an experiment to achieve validated learning.
Don't find the average consumer, find the early adopters. They are more forgiving of mistakes and eager to give feedback.
"Success is not delivering a feature; success is learning how to solve the customer's problem." - Mark Cook
Steer
Minimize time through the build, measure, learn loop.
Minimum Viable Product (MVP): version of the product that enables a full turn of the Build-Measure-Learn loop with a minimum amount of effort.
Must measure the impact of an MVP. Get it in front of potential customers.
Planning for BML loop actually happens in reverse: figure out what we need to learn, figure out what we need to measure to know if we are learning, then figure out what we need to build to run that experiment.
Leap
genchi gembutsu: "go see for yourself."
First step, confirm your leap-of-faith questions, confirm the customer has a problem worth solving.
If we do not know who the customer is, we do not know what quality is.
craft a customer archetype, define your target customer
Test
Early adopters don't care about faults in the product, only that they are the first to use a new technology.
When in doubt, simplify your MVP. Remove anything that does not contribute directly to the learning you seek.
The video MVP: just release a demo video, garner interest. (e.g. Dropbox)
The concierge MVP: Go out of your way to provide for early customers, even if it's inefficient. [See Paul Graham on why you should do things that don't scale]
Don't hide from customers for fear of someone else taking your idea. Hand a random CEO your idea on a golden platter and they won't even look at it.
If you don't want to harm long term image, start MVP under different brand name.
Measure
How innovation accounting works: use MVP to establish real data on companies current position, tune the engine towards the ideal, pivot or perservere.
Cohort analysis: look at each group of customers rather than cumulative totals. Provides much more insight into effect of changes made.
Vanity metrics: Metrics that look good but tell you nothing. Cumulative totals, etc.
Optimizing one part of the system undermines the system as a whole.
Split-test experiment (A/B testing): different versions of the product offered at the same time to different customers.
Incorporate split-testing into product development, not just marketing.
Three A's: Actionable, accessible, auditable. Reports must fulfill the Three A's
Only 5% of entrepreneurship is the big idea, the business model...95% is the gritty work measured by innovation accounting.
Pivot (Or Perservere)
Land of the living dead - successful enough to stay alive, but not growing like the founders/investors would like.
customer segment pivot - change audience focus
True measure of runway is how many pivots a starup has left, not just time. Encourages faster pivoting.
Why founders don't pivot faster: vanity metrics, unclear hypothesis==ambiguous results, fear that their vision was wrong.
Sign a pivot should be considered: decreasing effectiveness of product experiments, feeling that product dev should be more productive.
Have a regular 'pivot or perservere' meeting.
Batch
Work in smaller batches - quality problems can be identified much sooner.
Engineers and designers should work together side by side on one feature at a time.
Continuous Deployment
Reduced batch sized means faster learning
The instinct to work in large batches is so strong that when a large-batch system malfunctions we blame ourselves rather than the large batch.
Large batches grow over time
Toyota's Just-in-time production method: instead of the warehouse trying to guess how many replacement bumpers to make, the stores keep a small stock and 'pull' a new bumper when someone buys one.
The Lean Startup technique converts push methods to pull, and reduces batch size.
Grow
New customers come from the actions of past customers.
- Word of mouth
- Side effect of product usage
- Funded advertising
- Repeat purchase or use
Three Engines of Growth
- Sticky engine of growth - retain customers long term
- Viral engine of growth - get customers to spread the word to their peers
- Paid engine of growth - use ads. Make sure Cost Per Acquisition < Lifetime Value
Focus on one engine at a time.
Adapt
Don't add bureaucracy just to look professional
Split-the-difference approach to negotiating leads to groups taking extreme positions, an arms race
adaptive organization - one that automatically adjusts its process to current conditions
If you have a problem, ask 'why?' 5 times. (like an annoying kid)
Fix the problem at each level of 'why?' (in a proportional way)
Make sure everyone is in the room for 'five whys' or it will turn into 'five blames'
upper management should say, "if a mistake was made, shame on us for making it so easy to make that mistake"
- Be tolerant of all mistakes the first time
- Never allow the same mistake to be made twice
Five whys will force you to confront unpleasant truths
When starting to use five why process, start with a small problem, to teach the team how to use it properly.
Appoint a five whys master as a moderator for each meeting
Innovate
Startup teams require 3 attributes:
- scarce but secure resources
- independent authority to develop their business
- personal stake in the outcome
Don't hide the innovation team, managers in the parent organization might feel betrayed
Make an innovation sandbox
Switching to validated learning feels worse at first.
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
Avoid breeding a new pseudoscience around pivtos, MVPs, and the like.
Lean startup must avoid doctrines and rigid ideology
=====
Also Recommended
Buy The Lean Startup on Amazon or check it out at your library.
Want to know when I post updates? Join my newsletter by sending an email to james@jamesstuber.com
Member discussion