16 Power BI Service Overview

From Desktop to the Cloud

Power BI Desktop is where reports are built. The Power BI Service (accessible at app.powerbi.com) is where they live, where they are shared, where data refreshes automatically, and where organizations collaborate around data. Understanding the Service is essential for anyone who wants their work to reach others, stay current, and scale beyond a single analyst’s machine.

This chapter introduces the Power BI Service, walks through its interface, explains how workspaces and datasets are organized, and covers the two features that make cloud-hosted reports genuinely useful in practice: scheduled data refresh and the Power BI Mobile app.

flowchart LR
    A[Power BI Desktop <br> Build & Transform] -->|Publish| B[Power BI Service <br> app.powerbi.com]
    B --> C[Workspaces <br> Organize content]
    B --> D[Datasets <br> Scheduled refresh]
    B --> E[Reports & Dashboards <br> Share & collaborate]
    B --> F[Dataflows <br> Centralized prep]
    B --> G[Mobile App <br> Consume on the go]
    classDef default fill:#2e4057,color:#ffffff,stroke:#ff9933,stroke-width:3px,rx:10px,ry:10px;

16.1 Understanding the Power BI Service

16.1.1 Power BI Service vs. Power BI Desktop

Two Tools, One Ecosystem

Power BI Desktop and the Power BI Service are complementary tools that serve different roles in the Power BI workflow. Desktop is a Windows application installed on a local machine. The Service is a cloud-based platform accessible from any browser on any device. Neither replaces the other. Understanding what each tool does and does not do prevents frustration and ensures you are using the right tool for each task.

Side-by-Side Comparison

Capability	Power BI Desktop	Power BI Service
Connect to data sources	Yes, full range	Limited (some sources only)
Power Query transformation	Yes, full editor	Dataflows only
Data modeling	Yes, full model view	Limited editing
Create reports and visuals	Yes, primary tool	Basic editing only
Write DAX measures	Yes	Yes (limited)
Publish to Service	Yes	N/A
Share reports with others	No (must publish first)	Yes
Schedule data refresh	No	Yes
Create dashboards	No	Yes
Manage workspaces	No	Yes
Configure Row-Level Security	Define roles	Assign users to roles
View on mobile	No	Yes (via Mobile app)
Works offline	Yes	No (requires internet)

[Insert screenshot showing Power BI Desktop on the left (showing the Report view canvas with visuals being built) and the Power BI Service on the right (showing a published report in reading view in a browser), illustrating the two different environments side by side]

The Recommended Workflow

Build and refine your data model, transformations, and report layouts in Power BI Desktop. Publish to the Power BI Service to share, schedule refresh, and collaborate. Make further edits in Desktop and re-publish to update the Service version. Avoid making significant changes directly in the Service, as those changes can be overwritten the next time you publish from Desktop.

16.1.2 Navigating the Power BI Service Interface

The Power BI Service Home Page

When you sign in to the Power BI Service at app.powerbi.com, you land on the Home page. Home serves as your personal starting point, showing recently visited reports and dashboards, recommended content from your organization, and quick access links to the items you work with most frequently. It is personalized to your activity and updates automatically as you open and interact with content.

[Insert screenshot of the Power BI Service Home page showing the recent items section, recommended content cards, and the left navigation panel with icons for Home, Create, Browse, OneLake data hub, Apps, and Workspaces]

The Left Navigation Panel

The left side of the Power BI Service interface contains a vertical navigation panel that provides access to every section of the platform. The key sections are:

Home — your personalized starting page with recent and recommended content
Create — shortcuts for creating new reports, dashboards, dataflows, and datasets
Browse — a searchable view of all content you have access to across all workspaces, filtered by type (reports, dashboards, datasets, dataflows, apps)
OneLake data hub — a centralized catalog of datasets and dataflows available across your organization for report builders to connect to and reuse
Apps — all Power BI apps you have installed or that have been shared with you
Workspaces — a list of all workspaces you belong to, with the ability to expand each workspace to see its contents

[Insert screenshot of the left navigation panel in the Power BI Service showing the icons and labels for each section listed above, with the Workspaces section expanded to show two workspace names]

The Top Navigation Bar

The horizontal bar at the top of the Power BI Service provides:

Search — a global search bar (the magnifying glass icon) that searches across all reports, dashboards, datasets, workspaces, and apps you have access to. Searching by name, field name, or even report content is supported
Notifications — an alert bell that shows system notifications, data refresh status, and sharing notifications
Settings — a gear icon providing access to personal settings, admin settings (for admins), and the Power BI download center
Help — a question mark icon linking to Power BI documentation, the community forum, and support resources
Account — your profile picture or initials in the top-right corner, showing your signed-in account and licence information

[Insert screenshot of the Power BI Service top navigation bar with the Search bar, Notifications bell, Settings gear, Help icon, and Account icon all visible and labelled]

Reading View vs. Editing View

When you open a report in the Power BI Service, it opens in Reading view by default. In Reading view, you can interact with the report (use slicers, click visuals, drill down, apply filters) but you cannot change the report’s layout, add or remove visuals, or edit DAX formulas.

If you have edit permissions on the report, a Edit button appears in the top-right toolbar. Clicking it switches to Editing view, which provides a simplified version of the Power BI Desktop report canvas with the Visualizations and Data panes visible for making changes. Editing in the Service is suitable for minor adjustments. For significant changes to the data model or transformations, always use Desktop.

[Insert screenshot of a report in Reading view in the Power BI Service, with the Edit button visible in the top-right toolbar, and the Filters pane and visual interaction controls visible]

16.2 Organizing Content

16.2.1 Workspaces in the Service

My Workspace

My workspace is a private personal workspace that exists for every Power BI user. Content published to My workspace is visible only to you. It is useful for personal exploration, testing reports before sharing them, and storing datasets you are working on individually. Content in My workspace cannot be shared with others directly or packaged into an app.

[Insert screenshot of the Power BI Service left navigation panel with “My workspace” highlighted, and the My workspace content page showing a personal dataset and report listed]

Shared Workspaces

Shared workspaces (also simply called workspaces) are collaborative containers created for teams, departments, or projects. Multiple users can be members of a shared workspace simultaneously, each with a role (Admin, Member, Contributor, or Viewer) that determines what they can do within it.

Content published to a shared workspace is accessible to all members according to their role. Shared workspaces are where production reports live, where teams collaborate on datasets, and from where apps are created and distributed to broader audiences.

To create a new shared workspace:

Click Workspaces in the left navigation panel
Click Create a workspace at the bottom of the workspace list
Give the workspace a name and an optional description
Configure the workspace settings (licence mode, contact list) and click Save
Add members by clicking Access in the workspace toolbar and entering email addresses or security group names with their assigned roles

[Insert screenshot of the Create a workspace dialog showing the workspace name field, description field, and workspace settings section, alongside a second screenshot of the Access panel showing users being added with role dropdowns]

Workspace Content Types

A shared workspace can contain four types of items, each represented by a different icon in the content list:

Reports — interactive multi-page reports published from Power BI Desktop or created in the Service
Datasets — the data models and connection definitions that power reports. A single dataset can power multiple reports
Dashboards — single-page tile collections pinned from reports, used for at-a-glance monitoring
Dataflows — centralized data preparation definitions that produce reusable, refreshable tables (covered in the Dataflows section)

[Insert screenshot of a shared workspace content page showing a list of items with type icons (report, dataset, dashboard, dataflow) next to each item name, along with columns showing the item owner, last refresh date, and endorsement status]

Use Workspace Folders to Stay Organized

As workspaces grow, a flat list of reports and datasets becomes difficult to navigate. In current versions of the Power BI Service, you can create folders within a workspace to group related content. Right-click anywhere in the workspace content area and select New folder, give it a name, and drag items into it. Common folder structures include grouping by report type (Executive, Operational, Analytical), by project name, or by data domain (Sales, Finance, HR).

16.2.2 Datasets and Dataflows in the Service

Datasets in the Service

A dataset in the Power BI Service is the published version of the data model you built in Power BI Desktop. It contains the table structures, relationships, DAX measures, and connection information needed to query and display data. When you publish a .pbix file, the dataset and the report are uploaded as two separate items. The dataset is what connects to the source data and refreshes. The report is what displays the data visually.

A single dataset can power multiple reports. If several teams need to report on the same underlying data, they can all build their own reports connected to one shared dataset, rather than each maintaining their own copy of the data. This is called a shared dataset or live connection model and is a key pattern for enterprise-scale Power BI deployments.

[Insert screenshot of a workspace content list with a dataset item highlighted, alongside the dataset settings page showing the data source credentials section, the gateway connection, and the Refresh history panel]

Dataset Settings and Endorsement

Each dataset in the Service has a settings page (accessible by clicking the three-dot menu next to the dataset name and selecting Settings). Key settings include:

Data source credentials — where you enter the authentication details (username, password, or OAuth) for connecting to the data sources. These credentials are stored encrypted in the Service and used during scheduled refresh
Gateway connection — where you configure the on-premises data gateway if the dataset connects to local data sources
Scheduled refresh — where you configure automatic refresh (covered in detail in the next section)
Q&A — where you manage synonyms and field descriptions for the Q&A natural language feature
Endorsement — where Admins or designated certifiers can mark a dataset as Promoted (generally reliable) or Certified (officially validated by the organization). Endorsed datasets appear with a badge in the content list and in the OneLake data hub, signaling to other report builders that they are trustworthy sources

[Insert screenshot of a dataset Settings page showing the Scheduled refresh section collapsed and the Endorsement section showing the Promoted and Certified options with their descriptions]

16.2.3 Dataflows Overview

What Is a Dataflow?

A dataflow is a cloud-based data preparation tool in the Power BI Service that allows you to define Power Query transformations centrally, independent of any specific report or dataset. The result of a dataflow is a set of clean, transformed tables stored in Microsoft’s cloud storage (Azure Data Lake Storage) that any dataset or report in the organization can connect to and reuse.

Think of a dataflow as a shared Power Query transformation pipeline that lives in the cloud. Instead of each report builder writing their own Power Query steps to clean and standardize the same source data, one person (or team) builds and maintains the dataflow, and everyone else connects their datasets to the dataflow’s output tables directly.

[Insert screenshot of the Power BI Service dataflow editor showing a Power Query-like interface with a list of queries on the left, a data preview in the centre, and Applied Steps on the right, alongside a connection to a source database at the top]

Why Dataflows Matter

Dataflows address one of the most common pain points in enterprise Power BI deployments: data preparation duplication. Without dataflows, ten different analysts connecting to the same source database will each write their own Power Query cleaning steps. When the source changes (a column is renamed, a new status code is added), all ten analysts must update their own files individually.

With a dataflow, the cleaning and transformation logic is written once in the dataflow. All ten analysts connect their datasets to the dataflow output. When the source changes, only the dataflow needs to be updated. Every downstream dataset picks up the change automatically on the next refresh.

[Insert screenshot showing the dataflow architecture: a source database on the left feeding into a single dataflow in the centre, with multiple Power BI datasets and reports on the right all connected to the dataflow output tables, illustrating the single-source-of-truth pattern]

How to Create a Dataflow

Navigate to a shared workspace in the Power BI Service
Click New and select Dataflow
Choose Add new tables to open the Power Query-like dataflow editor in the browser
Connect to a data source (the same sources available in Power BI Desktop are available here)
Apply transformations using the same Power Query interface familiar from Desktop
Click Save & close to save the dataflow definition
Configure a Scheduled refresh for the dataflow so its tables stay current
In Power BI Desktop, connect to the dataflow’s output tables using Get data → Power BI dataflows and build your dataset on top of the pre-cleaned data

[Insert screenshot of the “New” dropdown in a Power BI Service workspace showing “Dataflow” as one of the creation options, alongside the dataflow editor’s source selection screen]

Use Dataflows for Organization-Wide Dimension Tables

The most impactful use of dataflows is for shared dimension tables that multiple reports across the organization need: a standardized Customer table, a canonical Product hierarchy, a consistent Date table, or a clean Employee roster. Build and maintain these once in a dataflow, then have every report builder connect their datasets to the dataflow versions rather than building their own from scratch. This ensures that everyone is working with the same customer names, the same product categories, and the same date definitions across all reports.

16.3 Keeping Data Current

16.3.1 Scheduled Data Refresh

Why Scheduled Refresh Matters

When you publish a report from Power BI Desktop, the dataset uploaded to the Service contains a snapshot of the data at the time of publishing. Without scheduled refresh, that data never updates. Every chart, every KPI, every trend line would show the same numbers it showed on the day it was published, growing staler every day.

Scheduled refresh configures the Power BI Service to automatically reconnect to your data sources at defined intervals, re-run all Power Query transformations, reload the refreshed data into the dataset, and make the updated data immediately available to all reports built on that dataset. From the consumer’s perspective, the report simply shows current data every time they open it.

[Insert screenshot of the dataset Scheduled refresh settings page showing the Refresh frequency set to Daily, the time set to 6:00 AM, the time zone configured, and a list of additional refresh times added below]

How to Configure Scheduled Refresh

In your workspace, find the dataset (not the report) in the content list
Click the three-dot More options menu next to the dataset name and select Settings
Expand the Scheduled refresh section
Toggle Keep your data up to date to On
Set the Refresh frequency: Daily or Weekly
Set the Time zone to match your organization’s local time
Set the Time for the refresh to run. Click Add another time to schedule multiple refresh runs per day (Power BI Pro supports up to 8 refreshes per day, Premium up to 48)
Toggle Send refresh failure notifications to On and enter email addresses to notify if a refresh fails
Click Apply to save the schedule

[Insert screenshot of the Scheduled refresh section expanded in the dataset Settings page, showing all the configuration options: the On/Off toggle, Refresh frequency dropdown, Time zone selector, refresh time inputs, and notification email field with the Apply button visible at the bottom]

Data Source Credentials

Before scheduled refresh can run, the Service must have valid credentials to connect to the data sources. These are configured in the Data source credentials section of the dataset Settings page:

Expand Data source credentials in the dataset Settings page
Click Edit credentials next to each data source listed
Enter the authentication details: OAuth, username and password, or API key depending on the source type
Click Sign in or Save

Once credentials are saved, the Service uses them automatically during every scheduled refresh without requiring you to sign in again, unless the credentials expire or are revoked.

[Insert screenshot of the Data source credentials section showing a list of connected data sources with “Edit credentials” links next to each one, and one credential dialog open showing username and password fields]

On-Premises Data Sources Require a Gateway

If your dataset connects to any data source that lives on a server within your organization’s network (SQL Server, Oracle, file shares, Excel files on a local machine), the Service cannot reach it directly through the internet. An on-premises data gateway must be installed on a machine within the network, registered with the Service, and configured as the connection bridge for that dataset. Without a gateway, scheduled refresh will fail for any on-premises source, and the dataset settings page will display a warning indicating that a gateway connection is required.

Schedule Refresh Before Business Hours Begin

Set your scheduled refresh to complete before your consumers start their working day, so they always open the report to fresh data. For a team that starts at 9:00 AM, scheduling the refresh at 6:00 AM gives a three-hour buffer that accommodates data source delays and any retry attempts if the first refresh fails. Add a second refresh at lunchtime if the data changes significantly during the morning.

16.4 Accessing Reports on Mobile

16.4.1 Power BI Mobile App

Consuming Power BI Content on the Go

The Power BI Mobile app is a free application available for iOS (iPhone and iPad) and Android devices. It allows users to access all the Power BI Service content they have permission to view from their mobile device: reports, dashboards, apps, and notifications. The mobile app is designed for consumption rather than creation. Report builders use Desktop and the Service. Report consumers use the Mobile app.

The app is particularly valuable for field teams, executives, and operational staff who need to check key metrics and dashboards while away from their desks. A sales manager in a customer meeting can pull up the current pipeline dashboard on their phone. A warehouse supervisor can check dispatch rates from the shop floor.

[Insert screenshot of the Power BI Mobile app on a smartphone showing a dashboard with several KPI tiles and a bar chart, demonstrating the mobile-optimized display with touch-friendly navigation]

Key Features of the Power BI Mobile App

The Power BI Mobile app provides the following core capabilities:

View reports and dashboards — all content shared with you in the Power BI Service is accessible in the app, with the same data and the same filter context as the browser version
Interactive filtering — slicers, filters, and cross-filtering work on touch screens with tap and pinch gestures
Notifications and data alerts — set alerts on dashboard tiles that send a push notification to your phone when a metric crosses a threshold (such as when daily revenue drops below a target or when a system metric exceeds a limit)
Annotate and share — capture a screenshot of a report page, annotate it with drawings or text using your finger, and share the annotated image directly from the app via email or messaging apps
Offline access — dashboards can be marked as favorites and are cached locally for viewing when no internet connection is available. Reports require a live connection
Barcode scanning — scan product barcodes to automatically filter a report to that specific product, useful for inventory and retail scenarios

[Insert screenshot of the Power BI Mobile app showing the Notifications/Alerts screen with a data alert configured for a revenue KPI, showing the threshold value and the notification that was triggered when the threshold was crossed]

Mobile-Optimized Report Layouts

Standard Power BI reports designed for desktop screens can be difficult to read on a small phone screen, as the layout is optimized for landscape wide canvases. Power BI Desktop provides a Mobile layout mode that lets report builders create a separate, portrait-optimized version of each report page specifically for mobile viewing.

To create a mobile layout:

In Power BI Desktop, go to the View ribbon and click Mobile layout
A phone-shaped canvas appears. Drag visuals from the panel on the right onto the phone canvas and arrange them in a single-column, vertically scrollable layout optimized for portrait orientation
Resize and reorder visuals to prioritize the most important metrics at the top
The mobile layout is published alongside the regular layout when you publish to the Service. The Mobile app automatically shows the mobile layout when available and falls back to the regular layout when it is not

[Insert screenshot of the Power BI Desktop Mobile layout view showing the phone-shaped canvas with a vertically stacked arrangement of three visuals (a KPI card at the top, a bar chart in the middle, and a slicer at the bottom), alongside the regular desktop layout of the same page visible in the background]

Prioritize Cards and KPIs in Mobile Layouts

Mobile consumers typically open the app for a quick status check rather than deep analytical exploration. Design mobile layouts with this behavior in mind: place the most critical KPI cards and status indicators at the top of the vertical canvas where they are visible without scrolling, followed by the most important chart or trend visual. Reserve the bottom of the layout for supporting details that the user can scroll to if they want more context. Keep the total number of visuals on a mobile page to five or fewer.

16.5 The Power BI Service at a Glance

16.5.1 The Power BI Service at a Glance

Everything Works Together

The Power BI Service is not a single feature but an interconnected platform where each component plays a specific role. Workspaces organize content and control access. Datasets store the data model and refresh automatically on schedule. Dataflows centralize transformation logic for reuse across multiple datasets. Reports and dashboards surface the data to consumers. The Mobile app extends access beyond the desktop. Together, these components form a complete data delivery platform that takes data from its source to decision-makers wherever they are.

Understanding how these components connect is what allows you to design a Power BI deployment that is not just technically functional but genuinely useful, maintainable, and trustworthy for the organization that depends on it.

The Service Is Where Power BI Becomes Organizational

A report that lives only in Power BI Desktop on one analyst’s laptop has limited impact. The moment it is published to the Service, scheduled to refresh automatically, shared with the right audience, and accessible on mobile devices, it becomes an organizational asset. That transition from personal tool to shared platform is what the Power BI Service enables, and it is the reason that learning the Service is just as important as learning how to build reports in Desktop.

Summary

Concept	Description
Service Building Blocks
Workspaces	Containers for collaboration and content lifecycle
Datasets	Reusable, governed semantic models
Reports	Authored visualisations connected to datasets
Dashboards	Pinned visuals that summarise key metrics
Apps	Packaged content distributed to users with a single link
Data and Capacity
Dataflows	Cloud-hosted Power Query producing shared transformations
Datamarts	Self-service warehouses with a managed SQL endpoint
Capacity Modes	Pro, Premium per User, Premium capacity, and Fabric