Power BI Guide: Everything You Need to Know to Get Started

Power BI is Microsoft’s reporting and analytics tool used for combining and visualizing data from various sources. Data can be imported both from an organization’s own systems and from cloud-based services and files. Power BI offers broad support for different databases, cloud services, and APIs, and support for new data sources is continuously expanding as the product evolves.

This Power BI guide was originally published in 2017 and has been regularly updated to reflect the current Power BI product. The latest update was completed by our experts Pentti Hämäläinen and Pauliina Heikkilä in May 2026.

Power BI is suitable both for fast, ad hoc self-service reporting and for more centralized, governed, organization-wide reporting driven by the business. As a result, usage patterns vary significantly, and reports—or broader reporting solution portfolios—can be created by a wide range of people, from end users to BI professionals.

Reports are created using the Power BI Desktop application or the Power BI Service cloud service. End users view reports in Power BI Service via web browser and mobile apps, or embedded, for example, in intranet solutions and Microsoft 365 services such as SharePoint, Teams, and PowerPoint, as well as in Power Platform solutions like Power Apps applications.

Reports can also be embedded into custom business applications by application developers, and embedding visualizations on public websites is also possible. Instead of the cloud service, reports can also be distributed using an on-premises Power BI Report Server.

You can explore publicly available Power BI reports in Microsoft’s Data Stories Gallery, which also offers great inspiration for your own reports and visual design choices.

Microsoft continues to actively develop Power BI and listens closely to user feedback. The development process is open: anyone can propose ideas, vote on them, and follow their progress in the official idea forum.

Microsoft Fabric is a comprehensive analytics solution for organizations. Fabric covers everything from data movement to real-time analytics and business intelligence. It offers a broad range of services—including data lakes, data processing, and integration—all from a single platform. It’s important to note that Fabric does not replace Power BI; rather, Power BI is an integral part of Fabric.

Fabric is Microsoft’s current way of delivering capacity-based Power BI capabilities. It replaces the previously used Power BI Premium capacity and includes the features that were part of Premium (excluding Power BI Report Server in smaller capacities). Some organizations may still continue to use Power BI Premium based on existing agreements.

Capacity is available in several different sizes (F2–F2048). This enables the use of premium-level features even in smaller environments compared to the previous Premium model. Fabric capacity can be used either as a continuous reservation or billed based on usage, and the capacity can also be paused and restarted as needed.

Report creators still require a Power BI Pro license. Report viewers are included in the Fabric capacity starting from the F64 level, allowing reports to be shared with a large audience without separate user-specific Pro licenses. In smaller capacities than this, report viewers also require a Power BI Pro license.

Roughly, the Fabric F64 tier corresponds to Power BI Premium P1. Its indicative price is approximately €4,500 per month as reserved capacity or about €10.40 per hour when billed based on usage. Smaller Fabric tiers start at around €140 per month or approximately €0.33 per hour with usage-based pricing (F2). You can check the latest pricing details on Microsoft’s website.

With Power BI, data can be presented through a wide range of metrics and interactive visualizations, making it an excellent tool for self-service BI. Report creation can be moved closer to end users—or even handled by the users themselves. Thanks to its ease of use and affordable pricing, Power BI has become a key reporting and analytics tool for roles such as controllers and analysts.

With minimal guidance, anyone can create new visual reports based on prebuilt data models. Building reports is easier and more intuitive than, for example, creating pivot tables in Excel. At best, a report page can be created in just a few minutes, and because visualizations are interactive, data can easily be explored from different perspectives and drilled into in multiple ways.

That said, it’s important to note that self-service BI can mean different things in different organizations, and the level of Power BI expertise required varies significantly depending on the chosen model.

1. Business-Led Self-Service BI

In smaller organizations, self-service BI may mean that a controller or finance manager builds reports entirely independently, without IT support. Data sources are often file-based, such as Excel files or CSV exports from various systems. In these cases, report authors typically cannot influence the structure of the source system data; instead, data is transformed into a reportable format using Power BI Desktop queries.

This approach requires strong Power BI Desktop skills—especially when combining data from multiple sources. When Power BI was first released in 2015, users were primarily independent self-service users, as the product initially did not support more centralized, IT-managed reporting.

2. IT-Managed Self-Service BI

In larger organizations, Power BI self-service reporting often means that users focus on creating reports and visualizations based on ready-made data models. IT centrally manages data warehouses, data models, and calculations, as well as their maintenance. This way, self-service users do not need to take care of data modeling or calculated measures, and the required level of technical expertise is significantly lower.

Self-service reporting can also be a combination of these two models, where reporting leverages both self-built data models and centralized models provided by IT.

Power BI’s role as a tool for centralized reporting has grown significantly, and in recent years it has increasingly been used as part of broader enterprise-level solutions. Microsoft Fabric has reinforced this direction by providing a unified platform where reporting, data models, and other analytics are more tightly integrated than before.

In many organizations, Power BI is now used in a way where IT is responsible for data structures, data models, and shared calculations, and Power BI serves as the reporting and visualization layer on top of these. In practice, reports are built on a centralized and governed data model, which may be backed by a data warehouse or another shared data source.

When reporting is used extensively across the organization, it is important to agree on common ground rules. Concrete decisions include, for example:

whether data is imported into Power BI and refreshed on a scheduled basis (Import) or

whether a direct connection to backend systems is used, allowing reports to display near real-time data (DirectQuery, Live Connection, Direct Lake).

The choice is typically influenced by the volume of data, performance requirements for reports (for example, how quickly reports are expected to respond to user interactions), and how up to date the data shown in the reports needs to be.

In practical implementations, centralized reporting is often not based on a single uniform model across the entire organization. For example, financial reporting may rely on a fully centralized data model, while more operational reporting makes use of more flexible solutions or a combination of different approaches. Power BI and Fabric enable these different models to coexist within the same environment, as long as the overall setup is managed in a planned and controlled manner.

Additional technical details can be found in Microsoft’s extensive white paper documentation.

Power BI users can typically be divided into four main roles:

1. BI developers, who are responsible for the backend reporting solutions, centralized data modeling, and shared calculations. They build and maintain the core structures of the Power BI environment as part of the IT or data team.

2. Technically oriented users, who also create Power BI solutions by writing queries, modeling data (such as defining table relationships), and building calculations and measures using DAX—working either in business teams or IT.

3. Report and dashboard authors, who understand the data well enough to create reports from existing models.

4. End users who utilize reports and dashboards to support decision-making, with the assistance of Power BI Copilot.

Queries, data modeling, calculation logic, and visualizations are primarily implemented using the Power BI Desktop application. Many of the same tasks can also be carried out extensively in the Power BI Service cloud service, where development can be done through a web browser, although not all Desktop features are yet available.

The finished reports are published either to Power BI Service or to the organization’s own Power BI Report Server reporting server, from where they are shared to end users.

Queries and data models are typically created by someone who has a solid understanding of the organization’s information systems and data content—such as a BI specialist, controller, or analyst.

Anyone who understands, for example, the basic principles behind Excel’s VLOOKUP function or relational databases can learn to build simple data models consisting of a few tables. If the report is based on a single dataset—such as a database view or a single CSV file—very little prior expertise is required.

Building more complex solutions requires a good understanding of relational databases and the fundamentals of dimensional modeling, as well as a willingness to learn the basics of the DAX calculation language.

Power Query

Data loading and transformation are handled in Power BI Desktop using Power Query. Power Query offers hundreds of transformation and data-shaping options, such as replacing decimal separators, splitting columns, replacing values, transposing data, or grouping data in different ways. The query editor provides extremely versatile data transformation capabilities.

Data models

Creating a data model means defining relationships between imported tables based on a unique identifier, such as a customer ID. Data modeling also includes tasks such as formatting columns, defining sort orders, and classifying fields—for example, as geographic data—so they can be visualized on maps.

DAX

In addition, the data model is extended with formulas written in DAX, known as measures or calculated columns. To ensure the solution is easy to maintain and extend—and supports calculations efficiently—the model should ideally be implemented as a dimensional star schema: Why data modeling is important in Power BI

Visualization and reporting

Visualization and reporting involve creating report files and pages. Visuals are easy to build, and report pages are automatically interactive. For example, clicking a single bar in a column chart immediately affects what is shown in the other visuals on the same report page.

In addition to Power BI’s built-in visuals, report files can include custom visuals, and application developers can create additional visuals to meet specific needs.

Power BI Desktop has its roots in Excel

Power BI Desktop often feels familiar to those who already know Excel’s Power Query, Power Pivot, and DAX calculations. Power BI is based on the same core ideas but brings together these functions—previously available in Excel as separate add-ins—into a single, unified tool.

In practice, the same tasks are carried out in Power BI Desktop, but in a clearer and more controlled way. Data is retrieved and transformed in Power Query, which is also available in Excel; on top of that, a data model and DAX calculations are built in a Power Pivot-style approach, and finally the data is presented as interactive reports and visualizations.

The same technical foundation is also used in Microsoft’s broader analytics solutions. For this reason, Power BI Desktop serves as a natural tool both for creating individual reports and as part of a centralized and governed reporting ecosystem.

Although Power BI Desktop has its roots in Excel, it has also drawn strong inspiration from other reporting tools such as Tableau and Qlik. That said, Gartner has ranked Microsoft ahead of Tableau and Qlik in its evaluations for several years now.

You can start familiarizing yourself with Power BI Desktop by downloading and installing it either from Microsoft’s website or via the Microsoft Store. Also, make use of these resources: Get started with Power BI Desktop, or watch the Getting Started with the Power BI Desktop video on the Microsoft Power BI YouTube channel.

If the cloud service is used as the report distribution method, models and/or reports created with Power BI Desktop are published to Power BI Service, where reports and data models are shared and accessed via a web browser or mobile applications. The service acts as the central hub for report lifecycle management and distribution.

In the cloud service, reports and their associated data models are placed in workspaces. A workspace serves as the place where reporting is maintained. Workspaces are typically intended for report creators and administrators, not for actual end users. Individual reports can be shared with end users from a workspace, or they can be packaged into broader collections, i.e. Power BI apps (Power BI App). Apps provide a governed way to distribute report collections to different user groups without giving users access to the development environment.

When using the cloud service, a single Power BI solution can optionally be split into two parts:

1. a technical layer (shared semantic models) that contains queries and the data model, and

2. separate visualization files that are connected to the technical datasets via live connections.

This approach allows implementation responsibilities to be distributed across different people and enables different teams to create tailored reporting solutions from centralized data models for different target audiences, for example.

In the cloud service, data refresh can be scheduled, access rights to reports can be managed, and reports can be used in a web browser, on mobile devices, and as part of Microsoft 365 services such as SharePoint and Teams.

In addition, the cloud service offers, among other things:

• report usage statistics

• alerts that notify about changes in specific values

• user-specific data filtering

• visibility into dependencies between reports and data models

Reports can also be created and edited in a web browser in Power BI Service, but more complex data modeling and calculation logic are still often implemented in Power BI Desktop.

In practice, Desktop and Service complement each other rather than replace one another.

The Power BI cloud service also offers a range of ready-made apps. These are prebuilt solutions provided by Microsoft or third parties, based on specific cloud services such as Microsoft’s own business applications. They enable reporting to be deployed quickly without having to build all reports from scratch. Some apps are free, while others require the purchase of a separate license.

The apps include a predefined data foundation, reports, and views, and their use typically requires a user account for the underlying source system. The apps operate mainly directly in the Power BI cloud service, and unfortunately their structure or the underlying data model cannot always be modified in Power BI Desktop.

However, for some apps related to Microsoft’s own products, Power BI Desktop files are also available. Based on these, reporting can be extended, calculations modified, and additional data sources combined. Examples include ready-made templates for Dynamics 365 solutions, which can be used as a starting point for custom reporting.

For more information about the cloud service: Tutorial: Get started creating in the Power BI service

Alongside Power BI cloud services, Power BI Report Server is available for organizations that want to distribute Power BI solutions without using the cloud, entirely within their own data center. Report Server offers a more limited feature set compared to the cloud service—for example, dashboards are not supported, and Power BI solutions cannot be split into separate technical and visualization layers.

When using Report Server, you must use a version of Power BI Desktop optimized for Report Server.

Learn more: What is Power BI Report Server?

Power BI Desktop is free and can be installed on any Windows workstation. However, it is not intended for report consumers. To share and consume reports, you need either Power BI cloud licenses or, alternatively, Power BI Report Server.

Power BI has a strong Excel heritage, as highlighted in the Desktop section. The current Power BI product was released in 2015 and was preceded by Power BI for Office 365, which was largely based on models and reports built in Excel. If you come across online instructions that refer to this older Power BI for Office 365 model, they no longer reflect the current version of Power BI.

Although Excel is no longer at the core of Power BI, it remains an important part of many reporting solutions. In connection with Power BI, Excel is used, for example, in the following ways:

• Excel files are used as data sources,

• Power Pivot data models built in Excel are imported into Power BI Desktop,

• Excel files can be viewed in the Power BI cloud service alongside other reports,

• Excel is used for reporting on top of data models published to Power BI, for example through Pivot reports.

In this way, Power BI and Excel complement each other in practical reporting work, even though their roles are clearly distinct.