Manage analytics, visualizations, and insights
Analytics surfaces and what each answers
Customer Service ships several analytics surfaces. Customer service historical analytics covers case volume, resolution times, SLA compliance, agent and queue performance, and topics. Omnichannel analytics covers session-level metrics: wait time, handle time, transfer rate, channel mix, and sentiment. Knowledge analytics tracks article usage, feedback, and defects. Customer Service Insights adds AI-generated topics that cluster cases into themes and surfaces emerging issues, alongside KPI cards and shareable dashboards.
Custom reporting is built on Power BI - standalone reports over Dataverse or the data lake, or Power BI visualizations embedded in model-driven dashboards. Remember the difference between system charts and views (record-level, lightweight, configured in the app designer) and dashboards (multi-component layouts, interactive or classic). Historical analytics data is retained for a limited window and refreshed on a schedule; it is not real-time operational data.
Traps: questions often describe a need for real-time individual case status - that is a view or a queue, not historical analytics. Another trap is assuming embedded Power BI tiles inherit app security; the Power BI dataset's own permissions apply.
Rule of thumb: pick the surface by the question being asked - 'why are cases rising?' is topics and historical analytics; 'how is my queue right now?' is a view or dashboard. Worked example: a manager wants to know which product line drives the most escalations, so use historical analytics broken down by subject or topic rather than a hand-filtered case view.