Search This Blog

Tuesday, December 13, 2011

Lean Start Up: #4 - Validated Learning

Validated Learning

a. The old waterfall method of building something (especially in software development) requires a problem that is known and a solution that is also known. Easy when building the next iteration of something already built. But, not good for starting something from scratch.
b. How to Achieve Failure: Building something that nobody wants and building it on time, under budget! Essentially we are executing a bad plan.
c. Edwards Deming: “The customer is the most important part of the production line.” All we should do is gauge whether the customer cares about it.
d. Agile methods have their origins in IT departments. The problem is known, but the solution is unknown and uncertain. But, what if we don’t know who the customer is or what s/he wants?
e. Steve Blank (a mentor of Ries) tells us to know our customer and use a process, from creating a prototype, to testing it, to validated learning, to innovative accounting.
f. Case Study: What Ries learned at previous company: Customers did not want to use his product to connect with existing friends, they wanted to connect to new people and make new friends. PIVOT—had to make a major move.
g. Lots of code written: 25,000 lines. All had to be tossed because it didn’t appeal to teens—target market. PIVOT…created an IM network instead that attached to existing system.
h. Could have discovered this if he had simply experimented with creating a single webpage…to see if people tried it! Nobody wanted it…so, he didn’t need page 2.

No comments:

Google Analytics