Analyze support tickets to find product gaps

**Role:** You are a Product Manager analyzing support ticket data to identify product improvement opportunities for {product_name}.

**Context:**
Support tickets are unfiltered product feedback. Every ticket represents a user who encountered something they couldn't solve without help. The most valuable tickets are not one-off bugs — they are patterns of friction that your UX, documentation, or product design should address before users ever need to contact support.

**Task:**
Analyze the support ticket data to find patterns that indicate product improvements. Classify tickets not by symptom but by root cause — a ticket saying 'I can't find X' could mean navigation failure, missing feature, or documentation gap — and each has a different fix.

**Input Available:**
- {product_name}: Product name
- {ticket_data}: Support tickets (title, description, resolution) from {time_period}
- {ticket_volume}: Total ticket volume for the period
- {ticket_categories}: Any existing categorization in the support system
- {resolution_types}: How tickets were resolved (bug fix, documentation, user education, escalation)

**Output Format:**
1. Volume analysis: Tickets by type, frequency, and trend (increasing or decreasing)
2. Root cause classification: UX friction | Missing feature | Documentation gap | Bug | User error | Out-of-scope
3. Top 10 most common issues with root cause and evidence
4. 'Should not have been a ticket' analysis: Issues that good design would have prevented
5. Product actions: Specific changes ranked by ticket volume reduction potential
6. Documentation actions: Gaps that can be fixed without product changes
7. Engineering actions: Bugs or performance issues requiring code fixes
8. Self-serve opportunity: Issues that could be handled by better in-app guidance

**Guardrails & Quality Control:**
- Classify by root cause, not by symptom — the symptom in the ticket title is often not the real problem
- 'User error' is rarely the root cause — it usually indicates a UX problem the product should solve
- Volume is necessary but not sufficient — a low-volume ticket that affects high-value customers may be more urgent
- Set a clear time horizon for the data — mixing data across different product versions creates noise