This document discusses lean startups and key ideas from thought leaders in the field. It introduces Alfred Sloan and William Durant, founders of modern corporations like GM. It contrasts waterfall and agile development models and highlights the work of Steve Blank and Eric Ries. Blank's big idea was that startups fail because no one wants their products, so companies should "get out of the building." Ries introduced the concept of a structured approach using customer development and agile principles in a build-measure-learn loop. His key ideas were to minimize time in the learning loop and measure success through validated learning rather than time or money spent. The foundation of lean startups is to avoid building things no one wants and get through the
13. Steve Blank
BIG OBSERVATION:
Startups fail because
because nobody wants to
buy the products or
services.
BIG IDEA:
“Get out of the building!”
14.
15.
16. Eric Ries
BIG OBSERVATION:
Customer Development
plus Agile makes a big
Build-Measure-Learn Loop
BIG IDEA #1:
Structured approach to
addressing “startups”
19. Eric Ries
BIG OBSERVATION:
Runway is not time,
it’s iterations
BIG IDEAS #2 and #3:
“Minimize the total time
through the loop”
Unit of Measure is
“Validated Learning”
20.
21.
22.
23. Let’s Take a Step Back™
• Startups are situations of extreme uncertainty
• Different skills for those situations
• Big Ideas from Blank and Ries:
– Avoid building stuff that nobody wants
– Structured approach to addressing “startups”
– Get through the loop faster
– Unit of Measure is Validated Learning
Ya got that??
24. If you only learn one thing…
Foundation of the Lean Startup
Avoid building stuff nobody wants
and get through the loop faster
Everything else is just commentary.
27. Free Your Mind
(and the rest will follow)
Start with with a shift of
mindset from thinking
about delivering code as
an achievement to
thinking about a
measurable learning
as an achievement.
28.
29. Good Luck!
I live on the Interwebs at
http://scottporad.com
Hinweis der Redaktion
“Startup” is a situation which requires a certain set of management skillsBest illustrated by a story
Billy Durant vs. Alfred SloanDurant had the genius for product-market fit (even when the customer doesn’t know, like Ford and faster horses)Sloan had the genius for administrationNot a value judgment – different needs at different stages
7 early stage companiesSite at scottporad.comBest known for Cheezburger: 1 site to top 100 propertyDuring my time at Cheezburger that we really started to put words to some of the things we’ve been learning for the last decade
[Slide with stages]Started my career with thisBased on manufacturingStage-basedMeasure of progress was meeting milestonesSpec complete, code complete, etc.Works well for understood problems with known solutions
Agile’s BIG OBSERVATION: this isn’t working because we can’t define the solutionAgile’s BIG IDEA: shorter feedback loopsTook a lot from Lean Manufacturing and eliminating wasteMeasure of progress was working codeWorks well for understood problems with unknown solutions
Steve Blank, serial entrepreneurExplain the four steps of Customer Development
Eric RiesCD + Agile = Giant BML good or “early stage” [together slide]
John Boyd, USAF ColonelObserve – take in some inputsOrient – make a judgment Decide – form a plan of actionAct – do it (which generates more inputs)Pilots who could do this faster would win because they were a step ahead of their enemyOpposite of Gretsky – playing to where the puck wasOld vs. New PlanesChess: 2 or 3 moves to 1
In two ways:Often misconstruedit’s hard and takes time and effort
For example, it might be better for us to measure the number of A/B test results we have, than to measure the number of story points of code we deliver.Software development practices – agile, tdd, ci/cdCustomer development practices – split tests, mvps, interviewsMindset
identify a customer segment, establish a hypothesis about that segment, make a change to product or marketing, run an A/B test, and measure the impact of the change. With that new learning, then repeat the process again.