38 Power BI Service and Collaboration Features

38.1 Why Power BI Service Matters

Power BI Desktop is the workshop; Power BI Service is the marketplace.

Power BI Desktop is where the analyst builds. Power BI Service (the cloud-hosted SaaS at app.powerbi.com) is where the audience consumes. Without the Service, every Power BI report sits on the analyst’s laptop — no schedule, no sharing, no governance.

The standard reference for Power BI deployment patterns is Introducing Microsoft Power BI by Alberto Ferrari & Marco Russo (2016), which covers the Service alongside the Desktop. Chris Webb’s earlier work (Chris Webb, 2014) remains useful for understanding the underlying engine that powers cloud refreshes.

For a visualisation-focused book, this chapter is where the Power BI dashboards built in earlier chapters (Power Query in Chapter 35, DAX measures in Chapter 36, the column-vs-measure decision in Chapter 37) become shared institutional artefacts — published, scheduled, governed, and consumed across the firm.

38.2 The Power BI Ecosystem

flowchart TD
    PB["Power BI<br>Ecosystem"]
    PB --> D["Desktop<br>Authoring on<br>Windows workstations"]
    PB --> S["Service<br>Cloud SaaS at<br>app.powerbi.com"]
    PB --> M["Mobile<br>iOS, Android,<br>Windows apps"]
    PB --> RS["Report Server<br>On-premise SSRS-<br>compatible server"]
    PB --> E["Embedded<br>Power BI in custom<br>applications"]
    PB --> F["Microsoft Fabric<br>Unified analytics<br>platform with<br>Power BI built-in"]
    style PB fill:#e3f2fd,stroke:#1976D2
    style D fill:#fce4ec,stroke:#AD1457
    style S fill:#fff3e0,stroke:#EF6C00
    style M fill:#fff8e1,stroke:#F9A825
    style RS fill:#e8f5e9,stroke:#388E3C
    style E fill:#ede7f6,stroke:#4527A0
    style F fill:#f3e5f5,stroke:#6A1B9A

The Power BI Family at a Glance

Product	Role
Power BI Desktop	Free Windows authoring app for building reports and models
Power BI Service	Cloud SaaS for hosting, sharing, and refreshing reports
Power BI Mobile	iOS, Android, and Windows apps for consumers
Power BI Report Server	On-premise server for organisations that cannot use the cloud
Power BI Embedded	Embed Power BI dashboards in custom applications via Azure
Microsoft Fabric	The unified analytics platform with Power BI as one workload

For most modern deployments, Power BI Service (with Microsoft Fabric increasingly the umbrella) is the consumption platform. Report Server is the fallback for organisations with strict on-premise requirements; Embedded is for software vendors and customer-facing applications.

38.3 Workspaces

A Workspace is a collaborative container for Power BI content. Every Workspace holds reports, datasets, dataflows, dashboards, and apps; permissions are assigned at the Workspace level and inherit downward.

Workspace Roles

Role	Capability
Admin	Full control: add/remove members, manage settings, delete workspace
Member	Add and edit content, share, manage permissions on items
Contributor	Create and edit content; cannot manage workspace permissions
Viewer	Read-only consumption

The default place every analyst starts is My Workspace — a personal area not visible to anyone else. Production work belongs in a Shared Workspace with proper roles. A common pattern in mature organisations: one workspace per business unit (Marketing, Finance, Operations) with administrators in the data team and members in the business.

38.4 Licensing and Capacity

Power BI Licence Tiers

Tier	Per-User Cost	Capabilities
Power BI Free	Free	Personal use only; cannot share with other users
Power BI Pro	Monthly per user	Author and consume reports; share with other Pro users
Power BI Premium Per User (PPU)	Higher monthly per user	Premium features for individual users
Power BI Premium Capacity	Annual capacity-based	Dedicated capacity; share with Free users; paginated reports, larger models
Microsoft Fabric Capacity	Modern successor	Unified Fabric capacity covering Power BI plus other Fabric workloads

The licence model is consequential for sharing strategy. Pro requires every consumer to have a licence; Premium Capacity allows free users to consume paid content. Most large enterprises run on Premium or Fabric capacity.

38.5 Publishing from Desktop

To publish a .pbix file to the Service:

In Power BI Desktop, Home → Publish.
Sign in with the Power BI account.
Choose the destination Workspace.
Power BI uploads the file; the Service splits it into a dataset (the model and queries) and a report (the visuals on top of the dataset).
The Service responds with a link to the published report.

For governance, the same dataset can be reused by many reports. A common pattern in mature organisations: build a single certified dataset per business domain; build many reports on top of it.

38.6 Sharing and Collaboration

Sharing Options

Method	Description	Best For
Direct Share	Share a report or dashboard with named users via the Share button	Ad hoc sharing
Apps	Bundle a curated set of reports into a Power BI App for an audience	Organisation-wide rollout
Microsoft Teams Embed	Embed reports as a Tab in Teams channels	Team-level workflow integration
SharePoint Embed	Embed reports as a web part on SharePoint pages	Existing SharePoint-based intranets
PowerPoint Embed (Live)	Live, interactive Power BI visuals inside PowerPoint slides	Executive presentations
Subscriptions	Email snapshots on a schedule	Daily / weekly digests
Email Subscriptions to Apps	Subscriptions to App content	Scaled distribution
Public Publish to Web	Publish a report to the public internet (high caution)	Public dashboards only; never personal data

The mature pattern: build a Power BI App for each major audience, with all reports the audience needs. Users install the App once and receive every update automatically.

38.7 Power BI Apps

A Power BI App is a curated collection of reports and dashboards bundled for a specific audience. Apps are the recommended way to share content at scale within an organisation.

To publish an App:

In a Workspace, click Create App (or Update App).
Setup: Name, description, logo, theme colours.
Content: Select which reports and dashboards in the workspace to include.
Audience: Define which AD groups or users see the App; different audiences can see different content within the same App.
Publish App.

The audience installs the App from their Power BI home; the App appears as a navigable collection of reports. Updates published to the source workspace propagate to the App on the next App update.

A typical large-firm pattern: one App per business unit, each containing 5–15 reports for the relevant audience, updated on a defined cadence (weekly or as new content is ready).

38.8 Subscriptions and Comments

Subscriptions: A user clicks the bell icon on a report → choose a schedule and a target. The Service emails a snapshot at the chosen time. Multiple subscriptions per report (different schedules, different recipients) are supported.
Data-Driven Subscriptions: Premium feature — emails are filtered based on the recipient (each user receives only their region’s data, generated dynamically).
Comments: Users can add comments on a report or specific visual. The comments are visible to all users with access; useful for collaborative interpretation, “@” mentions notify the named user.
Sensitivity Labels: Microsoft Information Protection labels (Confidential, Restricted, Public) can be applied to reports and travel with exports — a downloaded PDF retains the label.

38.9 Refresh Schedules and Gateways

Published datasets must be refreshed to reflect updated source data. The Service supports:

Cloud-Native Sources (Azure SQL, Snowflake, BigQuery, SharePoint Online, Dynamics): Direct refresh from the cloud, no extra infrastructure.
On-Premise Sources (on-prem SQL Server, file shares, Excel on a network drive): Require an On-Premises Data Gateway — a small Microsoft-provided service running on a server inside the corporate network, which the Service uses to reach the data.
Schedules: Daily, weekly, or up to 8 times per day on Pro; up to 48 times per day on Premium.
Incremental Refresh: Premium feature for very large datasets that refresh only the new partition.

The On-Premises Data Gateway is the single most-asked Power BI Service operations question. Most enterprise deployments operate at least one Gateway, often clustered for high availability.

38.10 Governance and Admin

Power BI Service offers a tenant-level Admin Portal for governance:

Tenant Settings: Control who can publish, share, export, embed, etc.
Usage Metrics: See which reports are most viewed, by whom, with what device.
Audit Logs: Track every publish, share, export, and view for compliance.
Capacity Metrics: For Premium / Fabric tenants, monitor capacity utilisation.
Workspace Audit: Inventory of all workspaces, owners, and content age.
Sensitivity Label Policies: Mandatory labels on certain content categories.
Endorsement and Certification: Mark a dataset or report as Promoted or Certified (the Power BI equivalent of Tableau’s certified data sources).

A serious Power BI deployment has explicit governance: who can publish, where, with what data, under what review process. Without it, the Service becomes a content swamp.

38.11 Microsoft Fabric — The Modern Direction

Microsoft Fabric, announced in 2023 and now generally available, is Microsoft’s unified analytics platform. It bundles Power BI alongside data engineering (Spark notebooks), data warehousing, real-time analytics, data factory pipelines, and OneLake (unified storage) under one umbrella.

For Power BI users, Fabric offers:

OneLake — a single, federated data lake every Fabric workload reads from.
Direct Lake — a new connection mode that combines Import-mode performance with Live-connection freshness.
Dataflows Gen2 — enhanced Power Query Dataflows.
Lakehouses and Warehouses — Fabric-native analytical stores.

Fabric is the strategic direction for new Power BI deployments. Existing Pro/Premium tenants continue to operate as before, with Fabric features layered on top.

38.12 Common Pitfalls

My Workspace as Production: Personal workspace used as the production location; nobody else can access or maintain content.
Direct Share Sprawl: Hundreds of one-off shares with named users; impossible to audit or revoke at scale.
No Apps: Sharing scattered across direct shares instead of curated Apps; users cannot find content reliably.
Premium for Tiny Models: Buying Premium capacity for one small report; expensive when Pro would suffice.
Pro for Mass Audience: Trying to share with thousands of users on Pro licences; capacity (Premium / Fabric) is the right answer.
No On-Premises Gateway: Refresh fails because the data is on-prem and no Gateway is configured.
Stale Gateway: Gateway running on an unpatched server; refresh fails after a Microsoft update.
No Endorsement: Multiple workspaces with overlapping datasets; users do not know which is the trusted source.
No Usage Metrics Review: Half the published reports are unused; nobody trims them.
Sensitivity Labels Skipped: Confidential data shared without sensitivity labels; downstream exports are unprotected.
No Backup Strategy: A workspace deleted by mistake takes content with it; no recovery path.

38.13 Illustrative Cases

A Bank’s Power BI App for the Branch Network

A retail bank publishes a single Power BI App for branch managers, containing a region dashboard, a product-mix dashboard, and a customer-segmentation report. The App reaches 1,200 managers via Premium capacity. Updates flow weekly from the central analytics team; managers consume on the Mobile app during morning rounds.

A Manufacturing Group’s Gateway Cluster

A manufacturing firm has plants in five locations, each with on-premise ERP. A clustered On-Premises Data Gateway runs in two corporate data centres, refreshing extracts from each plant nightly. The clustered gateway provides high availability — if one node fails, the other refreshes the data without dashboard downtime.

38.14 Hands-On Exercise: Publishing, Sharing, and Refreshing in Power BI Service

Aim: Publish a Power BI workbook to the Service, organise it in a workspace, configure a refresh schedule, build a Power BI App, and set up a subscription.

Scenario: A Yuvijen Stores analyst takes the Power BI dashboards built in earlier chapters and deploys them through the full Service workflow.

Deliverable: A published workspace with a refreshing dataset, a published App, a subscription, and a one-page sharing reference.

38.14.1 Step 1 — Create the Workspace

Sign in to app.powerbi.com.
Workspaces → Create a workspace.
Name: Yuvijen Sales Analytics. Description: Sales and customer dashboards for Yuvijen Stores.
Set advanced options: licence mode (Pro for now; switch to Premium later if needed).
Save.
Access: add the analytics team as Members and the business owners as Admins.

38.14.2 Step 2 — Publish the Power BI Desktop Workbook

Open the .pbix file in Desktop.
Home → Publish.
Select the Yuvijen Sales Analytics workspace.
Power BI uploads; receives a confirmation with a link.
Open the link to confirm the report renders correctly.

The Service splits the file into a Dataset (model + queries) and a Report (visuals).

38.14.3 Step 3 — Configure the Refresh Schedule

In the workspace, click the dataset’s options menu → Schedule refresh.
Data source credentials: provide credentials for each source (database, file path).
If any source is on-premise, configure an On-Premises Data Gateway first.
Scheduled refresh → On.
Choose times (e.g., daily 6:00 AM IST and 6:00 PM IST) and time zone.
Failure notifications: enter the email of the dataset owner.
Save.

The dataset will now refresh automatically on the schedule.

38.14.4 Step 4 — Endorse the Dataset

Click the dataset → Settings → Endorsement.
Mark as Promoted (analyst-vetted) or Certified (data-team certified, usually requiring Admin approval).
Add a description of what the dataset is for.

Other workspaces can now find and connect to this dataset; the certification badge tells consumers it is trusted.

38.14.5 Step 5 — Build a Power BI App

In the workspace, click Create app.
Setup: Name Yuvijen Sales, add a logo and description.
Navigation: Add the report and any dashboards from the workspace.
Audience: Define the audience (e.g., AD group Sales-Branch-Managers); different audiences can see different sections.
Publish app.

Members of the audience receive the App in their Power BI home and on the Mobile app.

38.14.6 Step 6 — Configure a Subscription

Open the report in the Service.
Click Subscribe (the bell icon at the top).
New subscription: name Yuvijen Daily Sales Snapshot. Frequency: Daily, 8:00 AM IST.
Recipients: enter email addresses (or AD groups).
Save.

The Service emails a PNG of the dashboard each weekday morning.

38.14.7 Step 7 — Embed in Microsoft Teams

In the relevant Teams channel, click + to add a Tab.
Choose Power BI.
Select the report from Yuvijen Sales Analytics workspace.
The report renders inside Teams, fully interactive, respecting permissions.

The team can now refer to live numbers in their daily standups without leaving Teams.

38.14.8 Step 8 — Document the Sharing Plan

The Sharing Reference Sheet

Audience	Access Method	Cadence
Branch Managers (1,200)	Power BI App via Mobile	Live + push notifications
Sales Leadership (15)	Workspace + Daily Subscription	Daily 8:00 AM IST
Regional Heads (5)	Workspace, Custom view	Live
Operations Council (executive)	Embedded in Teams channel	Live
Audit (annual review)	Endorsed dataset access only	Quarterly

The reference sheet documents who sees what, how, and how often — the operational form of the sharing strategy.

38.14.9 Step 9 — Connect to the Visualisation Layer

The hands-on closes the Power BI authoring-to-consumption arc:

Chapter 35 prepared the data with Power Query.
Chapter 36 expressed the analysis in DAX.
Chapter 37 chose measures over calculated columns.
Chapter 33 built the dashboard.
This chapter publishes, schedules, and shares it.

A workbook that lives only on the analyst’s laptop reaches one user; the same workbook published, scheduled, and bundled into an App reaches an entire organisation. The Service is what turns visualisation work into institutional capability.

Files and Screen Recordings

Screenshots of the workspace setup, refresh schedule, App publication, subscription, and Teams embed will be embedded here.

Summary

Concept	Description
Foundations
Why Power BI Service Matters	Power BI Desktop is the workshop; Power BI Service is the marketplace
Power BI Family
Power BI Desktop	Free Windows authoring app for building reports and models
Power BI Service	Cloud SaaS at app.powerbi.com for hosting, sharing, and refreshing
Power BI Mobile	iOS, Android, and Windows apps for consumers
Power BI Report Server	On-premise SSRS-compatible server for organisations that cannot use the cloud
Power BI Embedded	Embed Power BI dashboards in custom applications via Azure
Microsoft Fabric	Unified analytics platform with Power BI as one workload
Workspaces
Workspace	Collaborative container for Power BI content with role-based permissions
My Workspace	Personal workspace not visible to anyone else; rarely the right place for production
Workspace Admin	Full control: add or remove members, manage settings, delete workspace
Workspace Member	Add and edit content, share, manage permissions on items
Workspace Contributor	Create and edit content but cannot manage workspace permissions
Workspace Viewer	Read-only consumption
Licensing and Capacity
Power BI Free	Personal use only; cannot share with other users
Power BI Pro	Author and consume reports; share with other Pro users
Premium Per User	Premium features for individual users at higher per-user cost
Premium Capacity	Annual capacity-based licensing; share with Free users; paginated reports and larger models
Fabric Capacity	Modern unified Fabric capacity covering Power BI plus other workloads
Publishing
Publish from Desktop	Home Publish; Service splits the file into a dataset and a report
Dataset and Report Split	Service splits the .pbix into a Dataset (model + queries) and a Report (visuals)
Sharing Mechanisms
Direct Share	Share a report or dashboard with named users via the Share button
Apps	Bundle a curated set of reports into a Power BI App for an audience
Microsoft Teams Embed	Embed reports as a Tab in Microsoft Teams channels
SharePoint Embed	Embed reports as a web part on SharePoint pages
PowerPoint Embed Live	Live interactive Power BI visuals inside PowerPoint slides
Email Subscription	Email snapshots on a schedule
Data-Driven Subscriptions	Premium feature: emails filtered based on recipient
Public Publish to Web	Publish a report to the public internet; never use for personal data
App Setup
Power BI App Setup	Name, logo, navigation, audience configured in the App publication wizard
App Audience	Define which AD groups or users see the App; different audiences can see different sections
Subscriptions and Comments
Comments	Users add comments on a report or specific visual with @ mentions
Sensitivity Labels	Microsoft Information Protection labels travel with exports
Refresh and Gateways
Cloud-Native Refresh	Direct refresh from cloud sources without extra infrastructure
On-Premises Data Gateway	Microsoft service running inside the corporate network for refreshing on-prem sources
Refresh Schedule	Daily, weekly, or up to 8 times per day on Pro; 48 on Premium
Incremental Refresh	Premium feature for very large datasets refreshing only new partition
Governance and Admin
Tenant Admin Portal	Tenant-level governance interface in the Service
Tenant Settings	Control who can publish, share, export, embed at tenant level
Usage Metrics	See which reports are most viewed, by whom, with what device
Audit Logs	Track every publish, share, export, and view for compliance
Capacity Metrics	Monitor capacity utilisation for Premium and Fabric tenants
Workspace Audit	Inventory of all workspaces, owners, and content age
Endorsement and Certification	Promoted (analyst-vetted) or Certified (data-team approved) badges
Microsoft Fabric
OneLake	Single federated data lake every Fabric workload reads from
Direct Lake	New connection mode combining Import performance with Live freshness
Dataflows Gen2	Enhanced Power Query Dataflows in Fabric
Fabric Lakehouse	Fabric-native analytical store
Common Pitfalls
My Workspace as Production	Pitfall of personal workspace used as production; nobody else can access or maintain
Direct Share Sprawl	Pitfall of hundreds of one-off shares with named users; impossible to audit or revoke
No Apps	Pitfall of sharing scattered across direct shares instead of curated Apps
Premium for Tiny Models	Pitfall of buying Premium capacity for one small report; expensive when Pro would suffice
Pro for Mass Audience	Pitfall of trying to share with thousands of users on Pro licences when capacity is the right answer
No On-Premises Gateway	Pitfall of refresh failing because data is on-prem and no Gateway is configured
Stale Gateway	Pitfall of Gateway running on unpatched server; refresh fails after Microsoft update
No Endorsement	Pitfall of multiple workspaces with overlapping datasets and no certification trail
No Usage Metrics Review	Pitfall of half the published reports being unused while nobody trims them
Sensitivity Labels Skipped	Pitfall of confidential data shared without sensitivity labels
No Backup Strategy	Pitfall of a workspace deleted by mistake taking content with it; no recovery path