Business Context

You support the operations leader for Meta's user support organization across Facebook, Instagram, and WhatsApp. The leader wants a weekly business review (WBR) that quickly shows whether support operations are healthy, where performance changed week over week, and which issues need escalation.

Metric Scenario

Last week, inbound support volume rose from 4.8M to 5.4M cases (+12.5%). Overall SLA attainment fell from 92% to 86%, median first-response time increased from 3.1 hours to 4.4 hours, and cost per resolved case increased from $1.84 to $2.02. Resolution rate stayed flat at 78%, but recontact rate worsened from 14% to 17%. The operations VP asks you to design the WBR so it separates signal from noise, highlights leading vs. lagging indicators, and makes it easy to identify whether the issue is driven by channel mix, region, queue type, vendor site, or tooling changes in Meta's internal support workflow systems.

Requirements

1. Define the 6-8 core KPIs you would include in the WBR and identify one primary north-star metric for operations health.
2. Explain how you would structure the review so leaders can move from top-line performance to root-cause diagnosis in under 30 minutes.
3. Show how you would decompose the week-over-week deterioration in SLA and cost per resolved case.
4. Identify which metrics are leading indicators versus lagging indicators.
5. Propose thresholds or alert logic for when a metric should be escalated in the WBR.
6. Recommend 3 concrete actions if the decline is concentrated in one region, one queue, and one vendor site.

Data Available

• support_cases: case_id, product, issue_type, created_ts, first_response_ts, resolved_ts, region, channel, priority, vendor_site
• agent_activity: agent_id, vendor_site, shift_hours, productive_hours, cases_handled, qa_score
• workflow_events: case_id, routing_queue, automation_flag, tool_latency_ms, transfer_count
• customer_feedback: case_id, csat_score, reopen_flag, recontact_within_7d
• staffing_plan: week, vendor_site, scheduled_fte, actual_fte, absenteeism_rate