# Publishing and Dashboards in Power BI Service

> Chapter 14 of the Courseiva PL-300 curriculum — https://courseiva.com/learn/pl-300/power-bi-service-and-dashboards

**Official objective:** 3.3 — Publish reports and create dashboards in the Power BI service

## Introduction

Publishing and dashboards in Power BI Service. This is the moment your hard work building reports finally reaches the people who need it, but only if you do it right. For the PL-300 exam, understanding the difference between a published report and a dashboard, and how to manage them in the cloud, is where many candidates lose marks because they assume they are the same thing.

## The Restaurant Menu Board Analogy

Have you ever wondered why a restaurant has both a printed menu and a specials board? The printed menu is your polished report, full of details you can study at your table. But the specials board — that big chalkboard near the entrance — is your dashboard. It shows the most important things right now, at a glance: today's soup, the fresh catch, the chef's recommendation. You don't need to read the whole menu to decide if you want clam chowder; the board gives you the critical information immediately.

Now, publishing a report is like sending that printed menu to a central spot where all the restaurant's guests (your colleagues) can pick it up. You, as the chef, prepare the data, format it nicely, and then hit 'publish' — which is like handing the menu over to the hostess stand where everyone can grab a copy. Then, to create a dashboard, you take the most popular or urgent items from that menu and write them on the specials board. You pin only the key visuals — the profit margin, the customer satisfaction score, the number of orders today — so anyone walking into the Power BI Service 'restaurant' can see the health of the business without reading every page. The dashboard is live, automatically updated whenever the underlying data changes, just as the specials board is rewritten every morning. The report is the full story; the dashboard is the headline. In Power BI Service, your published report is the menu, and your dashboard is that chalkboard at the door.

## Core explanation

In Power BI, 'publishing' simply means taking a report you have built in Power BI Desktop and uploading it into the Power BI Service — the cloud-based version of Power BI that runs in a web browser. Think of Power BI Desktop as your workshop where you build the model and the visuals, and Power BI Service as the showroom where everyone else can view it. When you click 'Publish' in Desktop, you choose a workspace (a container or folder in the cloud) and the report lands there, along with its underlying dataset (the imported or connected data it uses). That dataset becomes a shared data source that others can use to build their own reports, if you give them permission.

A dashboard, on the other hand, is a single-page collection of visuals that you build entirely in Power BI Service. It is not a report; it is a summary. To make a dashboard, you must first have a published report in a workspace. From that report, you 'pin' individual visuals — charts, graphs, cards showing a single number — to a new or existing dashboard. You can also pin live data from other sources, like Excel or SQL Server Reporting Services reports. A dashboard can combine visuals from many different reports, all on one page, giving a broad view of the business. It is designed for monitoring: you or your colleagues can glance at it to see key metrics like sales totals, percentage complete, or customer complaints, without digging into the full report.

The key difference is that a report has multiple pages and is interactive — you can filter, cross-filter, and drill down into the data. A dashboard is a single, non-interactive canvas. You cannot filter a dashboard directly; it simply shows the pinned visuals. If you click a visual on a dashboard, it will take you back to the original report where you can explore further. This is a common exam trap: the exam will ask what happens when you click a tile (a single visual on a dashboard) — the correct answer is that it opens the underlying report, not that it filters anything.

Workspaces are the organisational units in Power BI Service. Each workspace has a role for its members: Admin, Member, Contributor, or Viewer. When you publish a report, you need at least Contributor role in the target workspace to be allowed to publish. Once published, you can share the workspace with others by assigning them a role, or you can share individual reports or dashboards directly using the 'Share' button. Sharing a report grants the recipient access to view it in the service, but they cannot edit it unless you give them the right permissions.

Dashboards also support 'natural language queries' via the Q&A feature. If you enable it, users can type questions like 'total sales last month' into a box, and Power BI will display the answer using the data in the dashboard. This is powered by the underlying dataset. The exam might test whether Q&A works on a dashboard (yes, if enabled) and whether it works on a report (no, Q&A is dashboard-specific).

Refreshing data is another crucial piece. When you publish, the dataset keeps the same connection settings you configured in Desktop. If you used 'Import' mode, the data is a snapshot that needs to be refreshed periodically. You set up a refresh schedule in Power BI Service using a gateway if your data source is on-premises. If you used 'DirectQuery' mode, the data is always live, but performance may be slower. The exam will ask about when you need a gateway (on-premises data sources) versus when you do not (cloud data sources like Azure SQL Database).

Finally, dashboards can have alerts set on them. For example, you can set an alert on a card visual showing 'total revenue' so that you receive an email if revenue drops below a certain threshold. Alerts only work on dashboard tiles that use certain visual types (like single-number cards and gauge charts). This is another specific exam point: which tile types support alerts? Only those that show a single value or a range.

To organise your workspace, you can create apps. An app is a packaged collection of reports, dashboards, and datasets that you can distribute to a large audience. Publishing an app is like publishing a whole magazine — it packages multiple items into one easy-to-browse experience. For the exam, remember: apps are for broad distribution; workspaces are for collaboration.

## Real-world context

Imagine you work as a junior data analyst for a retail chain with 200 stores. You have built a report in Power BI Desktop that shows daily sales, inventory levels, and staff hours for each store. Your director wants the regional managers to see a quick summary every morning without having to navigate a full multi-page report. Here is exactly what you would do.

First, you save your work in Power BI Desktop and click the 'Publish' button in the Home ribbon. You select the workspace called 'Retail Operations' (which you created earlier in Power BI Service). Power BI Desktop uploads the .pbix file, which contains both the report and its dataset. You get a success message saying 'Publishing succeeded'. Now, you open a web browser, go to app.powerbi.com, sign in, and find your report in the 'Retail Operations' workspace. You click on it to verify all the pages look correct.

Next, you create a dashboard. You click 'New Dashboard' in the left-hand navigation pane, give it the name 'Morning Briefing', and click 'Create'. You then go back to your published report and open the page that shows a map of store locations with sales figures. On the map visual, you hover over it and click the 'Pin' icon (shaped like a pushpin). A dialog box appears asking where you want to pin it: you select 'Existing dashboard' and choose 'Morning Briefing'. You repeat this process for a few other key visuals: a card showing total company sales, a gauge showing the percentage of stores meeting their target, and a line chart of sales over the past 30 days. All of these now appear as tiles on your dashboard.

You click the 'Edit' button on the dashboard to rearrange the tiles so they look clean. You resize the map to be larger, put the sales card at the top left, and arrange the rest in a logical flow. You also add a 'text box' tile where you type 'Data as of [timestamp]' so viewers know the last refresh time.

Now, you need to share this dashboard with the regional managers. You click the 'Share' button at the top right of the dashboard. A side panel opens. You type in the email addresses of the three regional managers. By default, they will get an email link that gives them direct access to the dashboard. You untick the box that says 'Allow recipients to share the dashboard' because you do not want them forwarding it to others without your knowledge. You click 'Grant access' and Power BI Service sends the invitations. The managers can now open the dashboard, see the key numbers, and click on any tile to drill into the underlying report if they need more detail.

Finally, you set up a daily refresh schedule for the underlying dataset. In Power BI Service, you click on 'Datasets' in the workspace, find your dataset, click the 'Scheduled Refresh' icon, and configure it to refresh every morning at 6:00 AM. Because your data source is a SQL Server database inside your company network, you first need to install and configure a data gateway (a small piece of software on a server in your office that securely connects to the cloud). You set up the gateway connection once, and then Power BI uses it to refresh the data automatically. Every morning, the dashboard tiles update with the latest numbers, and the regional managers see fresh data without you lifting a finger. This entire workflow — publish, create dashboard, pin tiles, share, and schedule refresh — is the core of what a BI professional does daily.

## Exam focus

The PL-300 exam covers 'Publishing and Dashboards in Power BI Service' in three main areas: 1) the mechanics of publishing from Desktop and managing workspaces, 2) the creation and configuration of dashboards, and 3) sharing, permissions, and data refresh. The exam will not ask you to write code or perform complex configurations, but it will test your understanding of the exact steps and the differences between concepts.

Question types you will see:
- Scenario-based multiple choice: 'You have published a report to a workspace. A user needs to see the report but not edit it. Which role should you assign?' Answer: Viewer. The trap is that 'Member' or 'Contributor' sound correct but allow editing.
- Drag-and-drop sequencing: 'Order the steps to create a dashboard from a published report.' The correct order is: Publish report from Desktop > Open report in Service > Pin visual > Create dashboard name > Arrange tiles. The trap is including 'Create dashboard' after pinning instead of before.
- Multiple-select: 'Which two of the following are valid ways to share content in Power BI Service?' Options include: Share from workspace, share via app, share via email attachment (this is fake — you cannot attach a .pbix to an email), and share via a direct link from 'Share' button. Correct: Share via workspace and Share via direct link.
- Build a dashboard simulation: You might get a simulated environment where you must pin a visual from a given report to a new dashboard. Familiarity with the Service interface is critical.

Key exam traps to watch for:
- The difference between a dashboard and a report: beginners assume they are interchangeable. The exam will present a scenario that requires one or the other. For example, if a user needs to cross-filter and drill through data, that is a report, not a dashboard.
- Permissions: You must know the four workspace roles (Admin, Member, Contributor, Viewer) and which actions each allows. For instance, only Admins and Members can publish apps.
- Data refresh: You will be asked whether a gateway is needed for cloud data sources (no) versus on-premises sources (yes). The trap is thinking all refreshes need a gateway.
- Natural language Q&A: The exam loves to test that Q&A is available on dashboards but not on reports. A typical question: 'You want users to ask questions in plain English about the data. What should you create?' Answer: A dashboard with Q&A enabled.
- Alerts: Only single-value tiles (card, gauge, KPI) support alerts. A chart with multiple bars does not.
- App distribution: Publishing an app is the standard way to give access to a large group, not sharing the workspace directly. The exam will ask: 'You need to provide access to 500 users. What is the recommended method?' Answer: Publish an app.

Definitions you must memorise:
- Tile: A single visual pinned to a dashboard.
- Dashboard: A single-page collection of tiles from one or more reports.
- Workspace: A container for reports, dashboards, datasets, and apps.
- App: A packaged distribution of a workspace's content.
- Gateway: Software that connects Power BI Service to on-premises data sources for scheduled refresh.
- Pinning: The act of adding a visual from a report to a dashboard.
- Shared dataset: A dataset that multiple reports can use, created when you publish a report with 'Include in report' (newer) or by publishing separately. The exam may reference shared datasets as a way to reduce duplication.

## Step by step

1. **Publish Report to Workspace** — In Power BI Desktop, click 'Publish' on the Home tab, select a target workspace in Power BI Service, and confirm. This uploads your .pbix file including the report and its dataset into the cloud workspace. Without this step, there is nothing to pin to a dashboard.
2. **Create a New Dashboard** — In Power BI Service, in the left-hand navigation, click 'New Dashboard', give it a descriptive name (e.g., 'Executive Summary'), and click 'Create'. This creates an empty canvas where you will later place pinned visuals. You can create multiple dashboards from the same report.
3. **Pin Visuals from Report to Dashboard** — Open your published report, hover over the visual you want to pin (e.g., a chart, card, or map), and click the pin icon. Choose an existing dashboard or create a new one. The visual becomes a tile on that dashboard. You can pin visuals from multiple different reports onto the same dashboard.
4. **Arrange and Customise Dashboard Tiles** — On the dashboard, click 'Edit' to enter tile editing mode. Drag tiles to rearrange them, resize them using corner handles, and add text boxes or images using the 'Add Tile' button. This step ensures your dashboard is visually clear and logically organised for viewers.
5. **Configure Data Refresh Schedule** — If your dataset uses Import mode, navigate to the workspace, select 'Datasets', choose your dataset, and click 'Scheduled Refresh'. Set the frequency (e.g., daily at 6 AM) and configure the gateway if your data source is on-premises. Without this, dashboard tiles will show stale data.
6. **Share Dashboard or Publish App** — Click the 'Share' button on the dashboard to grant direct access to individuals. Alternatively, for large audiences, publish an app from the workspace by clicking 'Create App', selecting the content to include, and distributing the app link. This final step puts your work into the hands of decision-makers.

## Comparisons

### Dashboard vs Report

**Dashboard:**
- Single page, non-interactive canvas of pinned visuals
- Cannot filter or cross-filter directly
- Created and edited only in Power BI Service

**Report:**
- Multi-page, fully interactive with slicers and drill-through
- Supports cross-filtering between visuals on the same page
- Created and edited in Power BI Desktop (then published)

### Shared Dataset vs Local Dataset

**Shared Dataset:**
- Stored in Power BI Service, usable by multiple reports
- Refresh schedule set in the service, not in Desktop
- Promotes consistency across reports

**Local Dataset:**
- Embedded within a single .pbix file, not shared automatically
- Refresh is configured within Desktop (if Import mode) or in service after publish
- Duplicated if the same data is used in different reports

### App vs Workspace

**App:**
- Packaged content for read-only consumption by many users
- Distributed via a link or directly installed in Power BI
- Recipients cannot edit the content

**Workspace:**
- Collaboration space for creators with roles (Admin, Member, Contributor, Viewer)
- Content is editable by members with appropriate roles
- Not designed for direct distribution to end users

### On-Premises Data Gateway vs Cloud Data Source

**On-Premises Data Gateway:**
- Required for scheduled refresh of on-premises databases (e.g., SQL Server in your office)
- Must be installed and configured on a server in the local network
- Adds a security layer by acting as a bridge

**Cloud Data Source:**
- Does not require a gateway for refresh (e.g., Azure SQL, Dataverse)
- Power BI Service connects directly over the internet
- Usually faster refresh and simpler setup

### Q&A on Dashboard vs Filter on Report

**Q&A on Dashboard:**
- Uses natural language questions typed into a box
- Works only on dashboards when enabled
- Answers are displayed as auto-generated visuals

**Filter on Report:**
- Uses explicit slicers, filters, or filter pane
- Works only within reports, not on dashboards
- Gives precise control over what data appears

### Share Dashboard (Direct) vs Publish App

**Share Dashboard (Direct):**
- Shares a single dashboard to specific users via email or link
- Recipients can view the dashboard and its tiles
- Best for small groups or one-off sharing

**Publish App:**
- Packages multiple reports, dashboards, and datasets into one application
- Recipients can browse all content in a single interface
- Best for large groups (hundreds of users) or external distribution

## Diagram

_Flow diagram showing the relationship between Power BI Desktop, Service, dashboards, sharing, refresh, and alerts._

```mermaid
flowchart TD
    A[Power BI Desktop] -->|Publish| B[Power BI Service Workspace]
    B --> C[Published Report]
    C -->|Pin Visual| D[Dashboard]
    D -->|Share| E[End Users View Dashboard]
    D -->|Set Alert| F[Email Notification on Threshold]
    B -->|Schedule Refresh| G[Gateway -> On-Premises Data]
    B -->|Direct Refresh| H[Cloud Data Source]
    B -->|Create App| I[App Distributed to Large Audience]
    E -->|Click Tile| C
    D -->|Q&A| J[Natural Language Queries]
    style D fill:#bbf,stroke:#333,stroke-width:2px
    style B fill:#f9f,stroke:#333,stroke-width:2px
```

## Common misconceptions

- **Misconception:** A dashboard is just another name for a report page in the Power BI Service. **Reality:** A dashboard is a separate feature in Power BI Service that stores pinned visuals from one or more reports, and it cannot be edited or filtered like a report page. (This confusion persists because in many other tools the word 'dashboard' is used interchangeably with 'report'.)
- **Misconception:** You can create a dashboard directly in Power BI Desktop and then publish it. **Reality:** Dashboards can only be created in Power BI Service, not in Power BI Desktop. You must first publish a report, then pin visuals from it. (Since users build visuals in Desktop, they naturally assume the dashboard creation happens there too.)
- **Misconception:** When you share a dashboard, recipients automatically get access to the underlying dataset and can build new reports from it. **Reality:** Sharing a dashboard gives only view access to that dashboard. To build reports, you need to share the dataset separately or add users to the workspace. (The principle of least privilege is not obvious to beginners, so they assume sharing one thing shares everything linked to it.)
- **Misconception:** Scheduled refresh will work automatically as long as the dataset is published, even if the data source is behind a firewall. **Reality:** For on-premises data sources, you must install and configure an on-premises data gateway and set up the connection in Power BI Service before scheduled refresh works. (Users underestimate the security and connectivity challenges of cloud-to-on-premises communication.)
- **Misconception:** You can pin any visual from any report to a dashboard, regardless of whether the report uses a live connection or imported data. **Reality:** You can pin visuals from any report regardless of the data source mode (Import, DirectQuery, or Live Connection). The limitation is not about data source type but about the visual type — for example, you cannot pin slicers or some custom visuals. (Users often confuse data refresh settings with pinning permissions.)
- **Misconception:** Publishing a report always creates a new dataset; you cannot use an existing dataset from another report. **Reality:** You can publish a report using an existing shared dataset by creating your report in Desktop based on that dataset, or by publishing a report that uses a dataset already in the workspace; the dataset will not be duplicated. (Many beginners think every publish is a fresh upload, not realising datasets can be shared and reused.)

## Key takeaways

- Dashboards exist only in Power BI Service, not in Power BI Desktop, and are built by pinning visuals from published reports.
- A dashboard tile is a single visual pinned to a dashboard; clicking it opens the underlying report for deeper analysis.
- Workspace roles determine what a user can do: Viewer sees content, Contributor publishes and edits, Member adds and removes others, Admin manages everything.
- To share a dashboard with someone who does not have access to the workspace, use the Share button and grant direct access via email.
- Alerts can be set on dashboard tiles that display a single value (card, gauge, KPI) to send notifications when a threshold is crossed.
- An app packages multiple reports and dashboards into one distributable bundle, ideal for giving access to large numbers of users.
- On-premises data refresh requires a data gateway; cloud data sources like Azure SQL Database do not need one.
- Natural language Q&A is available only on dashboards, not on reports, when enabled by the dataset owner.

## FAQ

**Can I edit a dashboard in Power BI Desktop?**

No. Dashboards are a Power BI Service-only feature. You can only create and edit dashboards in the web service by pinning visuals from published reports.

**Do I need a gateway to schedule refresh on a cloud dataset like Azure SQL?**

No. Cloud data sources hosted in Azure or other major cloud providers can be refreshed directly by Power BI Service without a gateway. Gateways are only needed for on-premises data sources.

**If I share a dashboard, will the other person see the original report too?**

Not automatically. Sharing a dashboard gives view access to the dashboard tiles. If the recipient clicks a tile, it opens the underlying report, but they need explicit access to that report's workspace to view it. You should share the report separately if needed.

**How many visuals can I pin to a single dashboard?**

There is no hard limit, but for performance and readability, Microsoft recommends keeping a dashboard to around 9-12 tiles. The service may warn you if you add too many.

**Can I set an alert on a bar chart tile on a dashboard?**

No. Alerts only work on single-value tiles such as cards, gauge charts, and KPIs. Bar charts do not support alerts because they contain multiple data points.

**What is the difference between sharing a workspace and sharing an app?**

Sharing a workspace gives users direct role-based access to all content (reports, dashboards, datasets) inside it. Sharing an app packages selected content and gives read-only access without exposing the workspace structure. Apps are recommended for broad distribution.

## Check your understanding

1. **Where can you create a new dashboard in Power BI?**

   Answer: Only in Power BI Service, by clicking 'New Dashboard' and then pinning visuals from a published report.

2. **What happens when a user clicks on a tile (visual) on a dashboard?**

   Answer: The tile opens the underlying report page that contains the original visual, allowing the user to explore the data further.

3. **Which workspace role is required to publish a report from Power BI Desktop?**

   Answer: At minimum, the Contributor role (or higher) in the target workspace.

4. **Why would you need to install a data gateway before setting up a scheduled refresh?**

   Answer: A gateway is required when the data source is hosted on-premises (inside a company network) to securely connect Power BI Service to it.

5. **What is the difference between sharing a workspace and publishing an app?**

   Answer: Sharing a workspace gives role-based access (including editing) to selected users, while publishing an app provides read-only access to packaged content for a large audience.

---

Interactive version with quiz and diagrams: https://courseiva.com/learn/pl-300/power-bi-service-and-dashboards