Problem
Placeholder — state the real problem in one or two sentences. Example: Close Friends is a static list. Nobody updates it, so it slowly stops meaning anything.
Who this is for
Name the actual user, not "users." One sentence on the person and the moment they're in when this problem shows up for them.
Proposed solution
Walk through the change like you're pitching it to a skeptical engineer. What's different, screen by screen. Keep it to the smallest version that tests the idea — a real PRD doesn't try to fix everything at once.
Scope is a decision, not a constraint you ran out of time to fight.
Success metrics
Say exactly what you'd look at two weeks after launch, and what number would tell you it worked. Vague metrics are the fastest way to look like you've never shipped anything.
Open questions
List two or three things you genuinely don't know yet. This is where a real PM sounds different from a student — comfortable saying "I'd need to test this" instead of pretending certainty.