Business Context

FinFlow is a B2B payments platform used by sales, operations, and executive teams. Three months ago, the analytics team launched a new executive dashboard to track revenue, conversion, retention, and payment success, but leadership says it is "informative but not actionable."

Metric Scenario

The dashboard is viewed by 42 weekly active stakeholders across Sales, Product, Finance, and Operations. In the last 8 weeks, average weekly viewers fell from 38 to 24, median time on dashboard dropped from 11 minutes to 6 minutes, and only 18% of viewers clicked through to a supporting drill-down report. Meanwhile, the company missed its quarterly net revenue target by 6%, payment success rate fell from 97.8% to 96.9%, and trial-to-paid conversion declined from 14.2% to 12.5%. The VP of Analytics wants to know whether the dashboard is effective, how to define that effectiveness rigorously, and what metrics should be used to improve it.

Requirements

1. Define what an "effective dashboard" means for this business and identify one primary success metric plus supporting metrics.
2. Distinguish between dashboard usage metrics and business outcome metrics, and explain how they should be connected.
3. Propose a framework to diagnose whether low effectiveness comes from poor adoption, poor metric design, weak actionability, or missing trust in the data.
4. Recommend 3-5 changes to the dashboard and how you would measure whether those changes improved effectiveness.
5. Identify guardrail metrics to ensure optimization does not increase usage while reducing decision quality.

Data Available

• dashboard_events: viewer_id, team, dashboard_id, event_type, timestamp, widget_id, click_target
• report_drilldowns: viewer_id, source_widget, destination_report, timestamp
• decision_log: decision_id, owner_team, linked_dashboard_widget, decision_type, timestamp, status
• kpi_daily_snapshot: date, net_revenue, payment_success_rate, trial_to_paid_conversion, logo_churn_rate
• data_quality_incidents: incident_id, affected_metric, start_time, end_time, severity