Focus on feedback. For anything thing you deliver to the customer, make sure you understand how you will determine whether you are heading in the right direction, or not.
Work to understand how to observe and capture what the customer wants. The basis of this work is “Design Thinking”. Specific tools that can be used include, Story boards, Story maps, Empathy maps, and Personas.
Become proficient at capturing needs as User Stories in user voice format (“as a” … “I want” … “so that” ..) and acceptance criteria as Behavior Driven Development (BDD) for (“given” … “when” … then” …) so you can communicate clearly with both the Team and the customers.
Go to the “Gemba”. Gemba (現場) is a japanese term meaning “the real place.” For Product Owners, the use is that you have to actually go and see what customers in order to truly understand their needs.
Work hard to build trust with Teams, stakeholders, and customers. This means you must lead by example, develop your empathy for others, and be authentic in your interactions and meet your commitments with all parties.
Leverage decentralized decision making and trust (but verify) others to do their job. When your expectation is not met, be clear about your expectation.
Leverage, and expect, metrics that help you understand upcoming coming work, what is possible, and current projects. Tools include velocity history, cumulative flow diagrams (CFD), and burn-down charts.
Expect that you have to explain both the vision over an over again. You cannot over-communicate this! And it is important to reinforcement.
Step into your role as a leader. This means you will need to need to both be a continuous learner yourself, and be prepared to share your knowledge to coach and mentor others.
Ignore “sunk cost” (the amount of money you have already spent on something) as you determine what to do next. You cannot change the past, but you can decide what is best to do going forward - the best return on your investment.
Be conscious of the cost of Teams. A typical fully burdened agile Team in the US costs about $1.2M per year. This means you are making investment decisions of $50,000 every 2 week iteration. Are you getting return on that investment?