After reading so many blog posts and articles about the lean startup movement it was truly refreshing to read the book that started it: The Lean Startup by Eric Ries. This blog post is a short summary of the book explaining a few of the main points.
It’s the third book I’ve read «cover to cover» on my iPad mini. The first that is truly work-related. I have to admit I’ve fallen in love with reading on a pad.
The Lean Startup book is made up of three sections: Vision, Steer and Accelerate - The three steps of a lean startup. Every section gives examples of companies doing it both right and wrong, and these companies are not only startups, some are big, established companies.
What really hit me while reading this book was Validated learning, the build, measure, learn cycle and A/B testing.
Build, Measure, Learn
As a software engineer, in a social circle of entrepreneurs and innovators, I get a lot of pitches. People have an idea, a market that needs to be filled, and all they need is someone to do some programming. They need a application that can do X for Y - and they will pay with equity in the company. That’s when I start digging a bit deeper:
- How many potential customers have you talked to?
- Did you pitch the idea or ask them indirectly about the problem area?
After these two questions it becomes clear that they only have assumptions, not facts. That’s where Validated Learning comes in: What is the easiest and cheapest way of testing these assumptions?
Minimum Viable Product
If the idea sounds interesting I usually suggest that we can build an MVP: A simple site presenting the idea with a field to leave your e-mail for an exclusive invite. Then we do some advertising in social media groups / feeds - maybe some cheap advertising trough Facebook and AdWords. Word of mouth has proven to be the most effective advertising channel.
By logging every user-interaction on the MVP we get an idea of how solid the idea is:
- How many people close the site at once?
- How many click around, mark the text?
- Who actually leaves their e-mail and clicks submit?
- How many type in their e-mail but don’t click submit?
Using these data we can go from having assumptions to having data that proves there is a problem area - saving both money and time if the result turns out to be negative.
Combining the user-interaction data with actually talking with potential customers shows us if the idea is worth investing more time in.
Remember to not pitch your idea directly. Start as simple as possible by simply asking them how they go about their day. We want to hear their thoughts before we’ve injected our idea into their minds.
Pivot or Preserve
Using all our newfound data we now can decide: Should we continue as we have - or de we need to pivot?
Pivoting means to go in a totally different direction because we have discovered that our assumptions were wrong. Hopefully we have discovered something through our data gathering that we can pivot to, but some times the smartest thing might be to put the idea down and think up something else.
Vanity metrics vs Actionable metrics
One important thing when deciding to pivot or preserve is that we base this decision on actionable metrics, not vanity metrics. Vanity metrics can be things like: «New signups in a month» and «Conversations between users per month».
This doesn’t really tell us to much about how our product is doing, more actionable metrics would be «How many new signups start a conversation with another user» and «How many established users start a conversation with new users and get a response». We don’t care how many users we have, we care about how many profitable users we have.
Two other things mentioned in the book is A/B Testing and The five whys, I’ll write a blogpost about them at a later time.