5.3.3. Reports vs. Dashboards
💡 First Principle: The distinction between Reports and Dashboards is about depth versus breadth. Reports enable deep exploration of a single topic; Dashboards provide a surface-level overview of many topics. Using the wrong one frustrates users—an executive doesn't want to click through 20 pages; an analyst doesn't want a single-page summary.
| Feature | Report | Dashboard |
|---|---|---|
| Created in | Power BI Desktop | Power BI Service |
| Pages | Multiple | Single |
| Purpose | Deep analysis | Executive summary |
| Content | One dataset | Tiles from multiple reports |
| Interactivity | Full filtering, slicing | Limited (click to underlying report) |
⚠️ Exam Trap: This distinction is heavily tested! If the scenario asks for a "single-page overview" or "executive summary with tiles from multiple reports," the answer is Dashboard.
How the Pieces Fit Together
Reports and dashboards sit inside a larger hierarchy, and questions often turn on knowing which layer is being described:
| Layer | What it is |
|---|---|
| Workspace | The container. Holds everything below it and carries the access permissions |
| Semantic model (dataset) | The data itself — tables, relationships, measures. This is what visuals query |
| Report | Multi-page visuals built over one semantic model, authored in Desktop |
| Dashboard | A single page of tiles pinned from one or more reports, assembled in the Service |
| App | A packaged, read-only bundle of reports and dashboards published to an audience |
Pinning and Apps
A tile is a snapshot of a visual pinned from a report onto a dashboard. It refreshes with the data, and clicking it takes the viewer through to the report it came from — which is why a dashboard can summarise many reports without duplicating any of them.
An app is how you distribute at scale. Rather than sharing a dozen items individually, you package a workspace's content and publish it to an audience who get read-only access and simplified navigation.
⚠️ Exam Trap: dataset and dataflow are the pair most often confused. A dataflow prepares data — it runs Power Query in the Service and produces cleaned tables that can be reused by several models. A dataset (semantic model) is what reports actually query. If a scenario complains that the same preparation logic is duplicated across many reports, the answer is a dataflow; if it asks what the visuals are querying, that is the dataset.
