Business Context

AcmeCloud provides B2B workflow software and handles ~120,000 technical support tickets per month across chat, email, and phone. The Head of Support reports that the official ticket resolution rate improved from 78% to 91% in the last quarter, but customer complaints, repeat contacts, and account churn have also increased.

Metric Scenario

Leadership believes the current definition of a "resolved" issue is too operational and does not reflect whether the customer’s problem was actually fixed. Today, a ticket is marked resolved when an agent closes it or when the customer does not respond for 72 hours. Last month, 91% of tickets were marked resolved, but 18% of resolved tickets were reopened within 7 days, 22% of customers contacted support again about the same product area within 14 days, CSAT after resolved tickets fell from 4.5 to 4.1, and gross logo churn for accounts with 3+ support contacts was 2.3x the company average.

Requirements

1. Define a primary metric for whether a technical support issue is truly resolved, including the exact success criteria and time window.
2. Propose 3-5 supporting metrics or signals that distinguish superficial closure from durable resolution.
3. Explain how you would decompose the metric to identify where false resolutions are concentrated.
4. Identify likely trade-offs or failure modes if the support team optimizes only for closure speed.
5. Recommend what metric should be used for executive reporting versus frontline support operations.

Data Available

• tickets: ticket_id, account_id, user_id, created_at, closed_at, channel, priority, issue_category, product_area, close_reason, agent_id
• ticket_events: ticket_id, event_time, status_change, reopen_flag, escalation_flag, transfer_flag
• csat_surveys: ticket_id, survey_sent_at, score, response_text, responded_flag
• product_usage: account_id, user_id, event_time, feature_name, error_code, session_id
• account_health: account_id, plan_tier, renewal_date, expansion_flag, churn_flag, mrr