During the last ICP training one of the trainees asked me: “So, ok. You are telling us the cool stuff, unicorns and fairies. But in the real life we have to deal with a zillion of real-life-problems. For example we have several teams contributing to one product. To make an increment potentially shippable we need to perform integration testing which takes quite a time. Who should do that? When we should do that assuming that it cannot be done within a sprint? Who should perform that? What do your Scrum-fathers say?”

Great question.

Let me repeat my answer here.

  1. The need of having a potentially shippable product increment is not a whim of Ken Schwaber or Jeff Sutherland. It is a reliable and highly efficient tool to decrease time to market, increase predictability and enable the ability to manage the product scope. It was known long before Agile was born and will be there long after Agile dies. In complex environments with high turbulence increments are the vital tool to have in our business-survival kits. Full stop.
  2. Not many of us know that but this is how all the methods under Agile umbrella really work: You gather your team into the fist and do your best to deliver valuable piece of software in a short timeframe. And yes — you hit the walls. You may even start hitting them before the fist is formed. And you HAVE to remove the wall. Now. Not a year after, not someone else. You. The team. Now. And step by step removing the organizational walls (inability to create cross-functional teams, inability to deliver fast, inability to make the product of expected quality, inability to do what the users really need, lack of CI/CD, long process of server purchase, and — finally — inability to decide who is responsible for the integration testing and inability to perform it quickly). This is actually what makes you Agile and not the standups you are yawning at every day. This is actually what helps to do twice work in half time: you solving your problems just in time.
  3. And of course, the organizational changes and solutions seem to be very frightening. In order to soften that we have the short iterations and short experiments — if you feel uncomfortable, or it didn’t work or you failed because of the bad decision — roll back and think of another way.

A win-win for those who care and have the courage.