Power BI

    Power BI for figures everyone reads the same way.

    On Monday, the report shows Friday's sales, and sales and finance read the same figure. We first agree definitions and build the data model on top of them.

    For controlling, management and departments who regularly pull their figures together from several files and systems.

    • Power BI
    • Microsoft Fabric
    • Copilot in Power BI
    • Power BI Desktop
    • Microsoft Purview
    • Microsoft Teams
    The control hut of a ski lift in Davos, the chairlift in the snow outside, a Power BI report with bars and a donut chart on the laptop, the robot agent beside it
    01

    One decision, one report

    Power BI fits recurring analysis that is currently assembled from several files or systems. First we establish which decision the report supports and how recent its data needs to be. Later, the same model answers questions in Copilot and gives agents the agreed figures. For a one-off analysis, Excel is often enough; when data volumes and sources grow, Microsoft Fabric comes in.

    02

    Model, reports, Copilot

    We connect the agreed sources, prepare data and model measures, including from Power Apps, Dataverse and the ERP. Alongside presentation, we settle refresh, publishing and the upkeep of metric definitions. The building blocks:

    • Semantic modelTables, relationships and measures in DAX, with descriptions that people and Copilot understand.Result: every figure is calculated once and read the same way everywhere.
    • Reports and appsReport pages for the recurring decision, bundled in a Power BI app and embedded in Teams or SharePoint.Result: everyone finds the current report where they work.
    • Roles and accessRow-level security through Entra ID groups, workspaces by ownership, sensitivity labels from Purview.Result: each person sees the figures they are entitled to.
    • Data for CopilotAI data schema, verified answers and AI instructions on the model, then the "Approved for Copilot" setting.Result: Copilot answers from the figures you have agreed.
    • Refresh and deploymentA gateway for data from your own network, scheduled refresh with alerts on failure, deployment pipelines from development to production.Result: the data status is visible, and changes follow a fixed route.
    03

    Licences, capacity, data access

    Anyone who creates and shares reports needs Power BI Pro or Premium Per User; on a Fabric capacity of F64 or larger, people with a free licence and a viewer role can read along. Copilot requires a paid Fabric capacity of F2 or larger or Power BI Premium P1 or larger. We need data access, business definitions and reference figures for a known period, and we clarify gateway requirements and distribution to internal or external recipients.

    04

    Checked against reference figures

    The business team compares results with agreed reference figures. We test filters, roles with test users, failed refreshes and data-freshness indicators, and we ask Copilot the questions your team actually asks. Changes to sources or measures follow a documented process.

    Sound familiar? Tell us where you stand today.

    Contact

    What's new.

    As of: October 2026

    • As of 26 May 2026

      "Prep data for AI" prepares a model for Copilot.

      In Power BI Desktop and the service, you set which tables and fields Copilot uses, store verified answers to frequent questions and write AI instructions in your own terms. A model marked "Approved for Copilot" appears in standalone Copilot without additional warnings. The features are in preview.

      Source: Microsoft Learn: Prepare your data for AI
    • As of 24 August 2026

      Standalone Copilot searches across every report a person may see.

      The full-screen experience in the left navigation is in preview; the Copilot pane in reports is generally available. Usage is billed to a Fabric Copilot capacity or to a workspace on a capacity of F2 or larger; a Pro or Premium Per User licence alone is not enough.

      Source: Microsoft Learn: Copilot for Power BI overview
    • As of 15 June 2026

      From F64, people with a free licence can read reports.

      On Fabric capacities of F64 or larger, a free Fabric licence with a viewer role in the workspace is enough to read; below that, every reader needs Pro or Premium Per User. Microsoft is retiring the Power BI Premium P capacities and recommends Fabric capacities.

      Source: Microsoft Learn: Fabric licenses and capacity

    What controlling and IT ask before the first report.

    Pro, Premium Per User or a Fabric capacity?

    Pro is enough when everyone who creates or reads reports has a Pro licence. Premium Per User adds larger models and more refreshes per person. A Fabric capacity pays off when many people only read, from F64 with a free licence, or when Fabric workloads and Copilot are added. We work it out from your number of creators and readers.

    Does Copilot work in our tenant?

    Copilot needs a paid Fabric capacity of F2 or larger or Power BI Premium P1 or larger; Pro or Premium Per User alone is not enough. If the tenant or capacity is outside the United States and the EU data boundary, Copilot is off by default until the Fabric admin allows processing outside the region. We check your capacity's region and settings before the start.

    Can we connect data from our own network?

    Yes, through an on-premises data gateway that makes databases and files in your network reachable for scheduled refresh. We set it up on a dedicated server with a service account and define who is alerted when a refresh fails.

    When is Excel enough?

    For a one-off analysis or a report that one person maintains for themselves. Power BI pays off as soon as several people need the same figure regularly, data comes from several systems or roles should see different slices. Excel stays useful: it connects directly to the Power BI model.

    Can customers or partners see reports?

    Yes, as guests through Entra ID with their own workspace and row-level security, or embedded in your customer portal, where the application handles sign-in. Which route fits depends on the number of recipients and where they sign in today.

    Agree on a useful next step

    Tell us about your current workflow, the systems involved and the result you need. In our first conversation, we identify what else we need to know to plan a useful next step.