Wednesday, August 18, 2010

Lean startups

Through Shmula, I went to the Eric Ries' You Tube channel and I watched two of his interviews done by Scobleizer: The Lean Startup, Part I/II". Eric Ries writes and has popularised the term "Lean Startups.

At the beginning of his interview Eric establishes the software context: Compared to the traditional industries, this discipline has no physical constraints. Consequently the developers' imaginations are directly translated into a real product.
Eric has ten years experience as an entrepreneur and he saw so many talented people building software that nobody wanted. So he digs the concept of Lean start-up from his research to help entrepreneurs test their ideas right from the beginning of their activity.

At the beginning of a start up, Eric would:
  • Have a vision to change the world,
  • Test elements of his vision on customers,
  • Have a Problem Team, a part of the company dedicated to understand what customers' problem they try to solve,
  • Have a solution team, the other part of the company, who try to build the "minimum valuable product".

I do agree that the Lean idea of "minimum viable product" can easily be understandable in a physical world, due to its own constraints. Because of the flexibility, nothing forces developers to face their bad habits. Software guys tend to wait, to work and wait more, to feel their bag of new stuff is big enough to launched. Great enough from their point of view usually means a lot of stuff from their own perspective, only a little of which is valuable.

Eric gave a great example about a full product delivery up to a real customer, up to the far end of the production line. He gave this example about software refactoring oriented scalability, made available in production after 3 days of work. This Just in Time way is in opposition to large batch development.
Based on an example around employees training in a start-up , he drives us to the root cause, using the 5 whys principle. These 5 questions, most of the time, push an organisation to uncover layer by layer, the path to the real problem, through either technical or management structures.

In the second part, the discussion moves to agile principles, something which we understand much more nowadays. But he really emphasises the notion of testing your vision by steps, and to make sure the new features change the behaviour of the customer.

I really enjoy listening to Eric, because he has enough experience to step back and to give us valuable feedback. 

Sources :
Shmula : The Lean Startup, Video Interview, Eric Ries, Part 1
Youtube :StartupLessonsLearne Channel
Youtube : Scobleizer Channel : Lessons Learned : Lessons Learned: Minimum Viable Product: a guide

No comments :