[tools] The Lean Startup

[Tom Allen]

This is a bit strange as a tools email, but bear with me... I also wrote this a while ago, but then left it sitting in my drafts folder since I'd taken the book to work and accidentally left it there.

Sales Pitch: What is it, in 25 words or less?

The Lean Startup is a new approach to business that measures success in terms of how much the business has learned about the pathway to sustainable profit. It's also a book by Eric Reis, and a movement that is growing and being adopted around the world.

Who would use it?

Predominantly startup businesses, where startup can be reasonably well defined as "a human institution designed to create a new product or service under conditions of extreme uncertainty". However, this definition is broad, and a lot of what many of us on this list do day-to-day is entrepreneurial at least in intent. Some of the tools and techniques covered by the lean startup approach are directly applicable to programming, design, and other creative, team-based domains.

Okay, so how does it work?

There are five main principles:

1. Entrepreneurs are everywhere.
2. Entrepreneurship is management.
3. Validated learning.
4. Build-Measure-Learn.
5. Innovation accounting.

The first two are fairly obvious if you accept Reis' list of suggested features of an entrepreneur, someone who; loves creating new ventures or products, is adept at organisational politics, is prepared to take risks, and is skilled at forming cross-functional teams. Lots of us on this list have these characteristics, and the reason I'm writing this email is that the techniques in the Lean Startup can be applied in existing businesses too, enabling efficiencies that weren't previously possible.

3) Validated learning, describes a concept which can ultimately be summarised as "test your assumptions". The name is distinctly not just 'learning', it's validated learning because what you've learnt should be true thereafter (or at least true of the group/time/location on or at which you tested it). A key example used in the book is the idea of first learning what customers want, before building it, ideally avoiding all the wasted design and coding effort of building features that no one ever uses. This also distinguishes the whole lean startup concept from lean manufacturing, or Kaizen, or Agile development, etc. It's not about making your programming more effective, it's about making sure your building the right thing in the first place. (Perhaps a large part of why I like the concept is that my Ph.D. thesis was damn near the same idea but applied to online decision making and trajectory planning. Why bother spending ten minutes optimising your trajectory to save five seconds of travel time?)

4) The Build-Measure-Learn loop is the core of Reis' entire innovation accounting process, and works as follows:

1. Build a Minimum Viable Product to find out where the business is right now.
2. Propose hypotheses about how to tune the business towards its ideal, and conduct experiments to determine whether the hypotheses were valid. (In part, this is a process of finding the right metrics, then acting to positively perturb them.)
3. Finally, Pivot if the hypotheses do not lead to a sustainable business model, or persevere if they do.

If you've heard the phrases "minimum viable product" or "fail fast" before, this is where they came from. (And on that topic, it's not your business that is supposed to fail fast, rather your crappy hypotheses about how it will become sustainable.) An iteration through the loop should generate validated learning, which can then propel the business forward, and lead to new hypotheses to test.

5) Innovation accounting is a set of tools that help measure the rate of growth of validated learning, i.e.: how fast you're getting through the Build-Measure-Learn loop. In particular, he identifies three "engines of growth", and associated metrics that need to be measured in each:

1. "Sticky", where the growth in users is greater than the churn or attrition. i.e. new customer acquisition rate > churn.
2. "Viral", where the number of referred users who join for every user is greater than one. i.e. a "viral coefficient" > 1.
3. "Paid", where users are acquired by some paid technique (usually advertising) and then generate enough revenue to make the payment worthwhile. i.e. Cost of Acquisition (COA) < Lifetime Revenue (LTV, per user).

In Reis' words; "innovation accounting enables startups to prove objectively that they are learning how to grow a sustainable business." So determine your engine of growth, obtain tools that let you measure the relevant metrics, then test hypotheses and keep those ideas that measure up. Any metric that doesn't belong in the chosen engine of growth is probably a "vanity metric"; not only unnecessary to measure, but detrimental if its your focus. (How many of you have bosses measuring your contribution in lines of code? That's a vanity metric...)

Where can I get it?

http://www.theleanstartup.com

http://www.startuplessonslearned.com

Amazon, BookDepository, or if there's any left, one of those bricks 'n' mortar "bookstores".

What does it run on?

Entrepreneurs.

What are the alternatives?

There's a bunch of other good literature about this stuff, including Steve Blank's Four Steps to the Epiphany, Brant Cooper and Patrick Vlaskovits' The Entrepreneurs Guide to Customer Development, and a whole world of questions in the Quora Startups topic.

There's also the Silicon Beach mailing list and associated Sydney Meetup (Friday's 6pm, Bar 77, Grace Hotel, 77 York St, City).