The goals are to: articulate the problem clearly for everyone, distinguish between symptoms and root causes, define success criteria before brainstorming solutions, and identify constraints around the problem.

You'll get a well-articulated problem statement, clear boundaries defining the scope, and documented success metrics and constraints.

Be aware that people often want to jump to solutions. If someone says, 'The problem is we don't have feature X,' remind them that's a solution disguised as a problem. Ask: what user problem does feature X solve? Disagreement is valuable data. Don't immediately resolve conflicting problem framings. Explore the different perspectives. What evidence supports each framing? You might discover multiple problems. The 'Five Whys' technique can feel repetitive, but that's why it works. The initial answers are often symptoms, while the later ones get closer to the root cause. Encourage participants to push through the discomfort. Problems tend to expand in scope. 'Improve search' can easily become 'improve the entire information architecture.' Explicitly define the boundaries. You can always expand the scope later, but shrinking it is more challenging.