Lessons from Atlassian | Megan Cook (Head of Product, Jira)

Cook describes senior leaders sharing failures so experimentation feels less risky to discuss. The value is in what was learned, not in celebrating every failed result without judgment. A team should be able to examine a mistake rather than conceal it.

She prefers asynchronous updates when the purpose is only to report status. Meeting time is reserved for solving a problem. Clear documentation also lets people revisit why a decision was made and which assumptions supported it.

Her manager-report approach pairs short regular conversations with flexible calendar space for a deeper issue. A scheduled check-in cannot predict every difficult problem that will arise. The point is to make extra help accessible when it is needed.

Cook does not treat a polished final proposal as the whole approval process. She brings affected partners in early and returns as designs, data and tests develop. Their concerns can shape the work before a final decision rather than arriving as late surprises.

She wants a team to say what is known, what is expected and how the expectation will be tested. Treating a forecast as certainty makes it harder for others to add useful evidence. Being clear about uncertainty can support credibility rather than weaken it.

Cook recommends telling a group whether the request is for a decision, feedback or help with an uncertainty. A short narrative about the situation, change and implications gives context. Bring the key supporting data and be prepared to go deeper where needed.

Her internal-startup advice begins with the idea and solution that need to be proven, without adding people too early. The example comes from a large-company setting with existing planning processes. It is a reminder to match the process and team size to the maturity of the bet.