Are you developing something that has proven value?

If the value is proven, break the work into Pattern/Big Changes Come From A Lot of Small, Valuable Steps

If the value is not yet proven, Discover What Is Valuable Outside of the Building

• Split off the smallest team possible to test whether a new idea is valuable to the customer. This is the "discovery" team. Make sure they can

  • talk with the customer

  • design a "good enough" user experience

  • code a "good enough" front-end and and back-end proof-of-concept, and deploy it

• In parallel with the development team, begin a continous process where they

  • Take the next idea

  • Form a hypothesis: e.g.

    • we believe that if we provide feature then customers will _

  • Form a test card

    • for the test

      • Identify and/or recruit customers to be test subjects

      • Cut the idea in half

      • Cut it in half again

      • Design a "good enough" user experience

      • Code a "good enough" front-end and and back-end proof-of-concept

  • run the test: deploy. measure