Business Context
AcmeHome runs a 24/7 customer support operation across chat, email, and phone for its smart-home subscription product. The COO wants a weekly leadership report that consistently highlights operational trends, major events, and customer coverage issues before they affect renewals and NPS.
Metric Scenario
Last week, AcmeHome handled 182,000 inbound contacts, up 14% week over week. Overall SLA attainment fell from 92% to 84%, first-response time rose from 6.5 minutes to 11.2 minutes for chat, and unresolved backlog increased from 8,400 to 12,100 tickets. Weekend overnight coverage dropped to 71% of required staffing versus a target of 95%, while APAC email queues exceeded the 4-hour response target on 3 of 7 days. Leadership also noted two operational events: a mobile app outage on Tuesday that drove a 38% spike in contact volume, and a new agent onboarding class that reduced average handle time by 6% for simple billing cases but increased escalation rate from 9% to 13%.
Requirements
- Define the core KPI set you would report weekly to senior leadership, separating trend metrics, event metrics, and coverage metrics.
- Show how you would structure an executive dashboard so leaders can quickly identify whether issues are driven by demand, staffing, process quality, or channel mix.
- Explain how you would distinguish leading indicators from lagging indicators in the report.
- Identify the most likely drivers of last week's deterioration and how you would decompose them.
- Recommend what thresholds, alerts, and narrative commentary should be included in the recurring report.
Data Available
- contact_events fact table with contact_id, created_at, channel, region, issue_type, first_response_at, resolved_at, status
- staffing_schedule table with interval_start, team, region, planned_agents, staffed_agents
- incident_log table with incident_id, start_time, end_time, severity, affected_product_area
- agent_performance table with agent_id, date, handle_time, csat, escalation_rate, queue
- backlog_snapshot table with snapshot_date, queue, open_tickets, overdue_tickets
