Skip to content
Tutorials Center
My learning

Topics

Lesson 1 of 56 ยท Welcome to Power BI

Self-Service BI and Business Intelligence Concepts

Understand business intelligence and self-service BI: what BI is, its benefits and risks, the main tools, and how Power BI Desktop, Service, Mobile and Report Server fit.

Before you click anything in Power BI, it helps to know what problem it solves. Business intelligence (BI) is the practice of turning raw data into information that people use to decide things. Self-service BI (SSBI) is the idea that the people who ask the questions, such as a sales manager or a finance analyst, can answer many of them themselves, without waiting weeks for the IT department to build a report.

This lesson gives you the vocabulary: what BI is and how it works, what self-service changes, its advantages and its risks, the main tools on the market, and every component of the Power BI family. It makes the rest of this track easier to place. Power BI cannot run on this site, so the code here is a small Python stand-in that shows the most important risk of self-service: two people calculating "the same" number and getting different answers.

What is business intelligence?

BI is the combination of data, technology and habits that lets an organization measure what is going on. A typical BI process has five steps:

  1. Collect data from operational systems: sales, accounting, websites, spreadsheets.
  2. Integrate and clean it: fix types, remove duplicates, join sources (see Cleaning Data in Power Query).
  3. Store and model it in a structure that is easy to query, usually a star schema (Star Schema).
  4. Analyze it with calculations (measures) and exploration.
  5. Present it as reports and dashboards, and then act on what they show.

The questions BI answers fall into a few families: descriptive (what happened?), diagnostic (why?), predictive (what will happen?) and prescriptive (what should we do?). Traditional BI concentrates on the first two; Power BI also has forecasting and AI features for the others, as in The Analytics Pane.

Traditional BI versus self-service BI

In traditional (centralized) BI, a specialist team owns the whole chain. Business users submit a request, wait, receive a report and ask for a change when they have a new question. The results are consistent and governed, but the process is slow, and the backlog grows.

In self-service BI, business users connect to approved data, build their own visuals and reports, and explore. IT provides the platform, the data and the rules; users provide the questions.

Traditional BI Self-service BI
Who builds reports IT or a BI team Business users and analysts
Speed of a new question Days or weeks Minutes or hours
Consistency High, one team decides Needs governance to stay consistent
Skills needed by users Reading reports Basic modelling and a tool such as Power BI
Main risk Backlog and shadow spreadsheets Many versions of the truth

Most organizations settle on a hybrid: IT or a BI team builds and certifies shared semantic models (one definition of revenue, one customer table), and business users build reports on top of them.

Advantages of self-service BI

  • Speed: users answer their own follow-up questions.
  • Fit: the person who knows the business builds the view of it.
  • Lower load on IT, which can focus on data quality and platform work.
  • Adoption: people use tools they helped build.
  • Better questions: easy exploration leads to more questions, and more insight.

Risks and how to control them

  • Many versions of the truth. Two teams define "active customer" differently and meeting discussions turn into arguments about whose number is right.
  • Shadow IT. Departments copy data into personal files, which then go stale, leak or break.
  • Quality and performance. Untrained authors build slow models on bad joins and publish them widely.
  • Security gaps. Self-service makes sharing easy, including sharing something that should not be shared.
  • Maintenance debt. Reports nobody owns keep running, and nobody trusts them.

The answer is governance, not prohibition: shared and certified semantic models, row-level security, naming standards, a catalog of trusted datasets, training and a clear owner for each report. Governance and Deployment covers the Power BI features that support this.

Try it: the multiple-versions problem

Two analysts count "active customers" from the same orders. Without a shared definition, they report different numbers. With one shared measure, the question disappears.

Python Stand-in: two definitions, one shared measure

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The two analysts report 3 and 2 for the same business, because one counts customer C, whose only order was fully refunded, and the other does not. ("Last 90 days" and "last three months" are also different windows in most months, which would add a second disagreement.) In Power BI the fix is a measure in a shared semantic model that both reports use, with a description that states the definition.

Try it: a governance health check

A simple inventory of reports already exposes most self-service risks. This stand-in flags reports with no owner, a private copy of the data, or a stale refresh.

Python Stand-in: an inventory of reports and its risks

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Half of the inventory needs attention, which is typical. A catalog like this, kept by the BI team, is the first governance tool most organizations build.

The Power BI family

"Power BI" is not one program but a set of parts. They work together, and each has a job.

Component What it is Typical user
Power BI Desktop A free Windows application to connect to data, shape it (Power Query), model it and design reports. Report authors and modellers
Power BI service The cloud site (app.powerbi.com) where you publish, share, refresh on a schedule, secure and collaborate. Everyone
Power BI mobile apps iOS, Android and Windows apps to view reports and dashboards, with alerts. Readers on the move
Power BI Report Server An on-premises server for organizations that must keep reports in their own data center; it has its own release cycle and fewer features than the cloud service. Regulated or offline environments
Power BI Report Builder A desktop tool to design paginated reports: pixel-perfect, printable documents such as invoices. Operational reporting
Power BI Embedded A way for software companies to put Power BI reports inside their own apps. Developers
On-premises data gateway A small program that lets the cloud service reach data inside your network. IT

Excel has older BI add-ins that Power BI grew out of: Power Query (get and transform data), Power Pivot (the data model and DAX), Power View (interactive reports) and Power Map (now called 3D Maps). Today the first two live on in Power BI Desktop itself, while Power View is retired. The Q&A feature that answers typed questions is also part of Power BI. Microsoft also folds Power BI into Microsoft Fabric, a wider analytics platform, in which Power BI is the visualization and semantic-model layer. Product names change from time to time, so check Microsoft's documentation for the current list.

One path from data to decision

Desktop Connect, shape, model, design.
Service Publish, refresh, share, secure.
Mobile Read and get alerts anywhere.

Most reports travel from left to right.

1 / 2

BI serves different people at different time scales, and the same data is shaped differently for each:

Level Who Typical questions Typical output
Strategic Executives, board Which markets should we enter? Are we on track for the year? A few headline KPIs, long trends, targets
Tactical Department managers Which campaign worked? Is this region behind plan? Monthly dashboards with drill-down by team or product
Operational Staff on the shop floor, support, logistics What needs doing today? Which orders are late? Near real-time lists and alerts

Design for the level. A board does not want 40 slicers, and a warehouse supervisor does not want a five-year trend.

Two techniques sit behind many BI reports. Data mining searches large data sets for patterns, groups and unusual values (customers who behave alike, transactions that look like fraud); Power BI exposes simple forms of it through the key influencers and decomposition tree visuals and through Python or R (AI Visuals and Q&A). Optimization goes one step further and asks which choice gives the best result, for example the cheapest delivery route, and is normally done outside Power BI, with the results shown in a report.

Power BI and SSRS

SQL Server Reporting Services (SSRS) is Microsoft's older reporting server for paginated, pixel-perfect reports such as invoices and statements that are printed or sent as PDF on a schedule. Power BI is built for interactive exploration: readers click, filter and cross-highlight.

Power BI SSRS
Style Interactive dashboards and reports Fixed-layout, printable documents
Authoring Power BI Desktop Report Builder or Visual Studio
Data Imported model or DirectQuery Queries sent to the source each time
Delivery Cloud service, apps, mobile Report server, e-mail and file subscriptions
Best for Exploring and monitoring Operational paper-style output

Use Power BI when readers need to explore; use SSRS or paginated reports in the Power BI service (built with Power BI Report Builder, the successor to SSRS reports) when the exact layout and page breaks matter. Power BI Report Server can host both kinds on your own servers.

Power BI and other BI tools

Power BI is one of several leading tools. Besides the five below, other self-service platforms you will meet include Domo, ThoughtSpot (search-driven analytics), TIBCO Spotfire, Zoho Analytics, Cognos Analytics, Yellowfin and Alteryx (strongest in data preparation). A fair way to compare them is by job, not by brand:

Tool Notes
Power BI Strong value, deep Microsoft and Excel integration, a full modelling layer with DAX, large community.
Tableau Highly regarded visual exploration and design flexibility; a different license model and a more visual-first modelling approach.
Qlik Sense An associative engine that highlights related and unrelated data as you click.
Looker and Looker Studio Looker defines metrics in code (LookML) for governed analytics; Looker Studio is a free, lighter dashboard tool for Google data.
Excel Familiar and flexible, but built for individual work, not for shared governed models.

All of them can build the same basic charts. The choice usually depends on the data platform you already use, the skills of your team, the price, and how much governance you need. Learning the concepts in this track (star schema, measures, filter context, refresh and security) transfers to any of them.

What this means for you as a learner

Being "self-service" does not mean skipping design. A good self-service author:

  • starts from a certified or shared semantic model when one exists, instead of building a private copy of the data;
  • writes measures with clear names and descriptions;
  • follows a star schema so the model stays fast and understandable;
  • documents the source and refresh of every report;
  • asks "who owns this when I am on holiday?" before sharing.

The next lesson, What Is Power BI?, looks at the three parts of the product and the vocabulary of semantic models, reports and dashboards in more detail.

Check your understanding

Quick check ยท 5 questions

of 5 answered You scored / 5 ยท perfect!

  1. 1. What is the main idea of self-service BI?

    Self-service BI shifts report creation to the people who ask the questions, on top of data and rules provided by IT.
  2. 2. Which is the biggest risk of uncontrolled self-service BI?

    Without shared definitions, the same metric appears with different numbers in different reports, and trust is lost.
  3. 3. Which Power BI component lets you design pixel-perfect printable documents?

    Report Builder creates paginated reports, such as invoices and statements.
  4. 4. Where do you publish, schedule refresh and share reports with others?

    The service is the cloud site for publishing, refreshing, sharing and securing content.
  5. 5. How do organizations usually control self-service BI?

    Governance keeps the speed of self-service while keeping definitions and data protection consistent.

Summary

  • BI turns data into decisions through collection, integration, modelling, analysis and presentation.
  • Self-service BI lets business users answer their own questions on governed data: faster, but it needs governance to avoid many versions of the truth, shadow IT and security gaps.
  • Power BI is a family: Desktop (author), service (publish, share, refresh), mobile, Report Server, Report Builder, Embedded and the gateway.
  • Power BI sits next to Tableau, Qlik, Looker and Excel; the concepts you learn here transfer between them.

Next: What Is Power BI?.