Getting Started

This page explains how to get started with Power BI and also how to navigate this site and which pages will be of value to you based on your experience level.

Getting Started - with Power BI

What do you want to use Power BI for?

There are three main use cases for Power BI:

  1. Visualization only; Creating visualizations on prepared data where no data transformation or modelling is required
    1. This can be the case when the source you are using is a cube (Microsoft SSAS) or a single table (excel, flatfile or any other source system)
    2. Visualizations in Power BI are very powerful and this mode still allows for quite some customization (adding filters, adding simple calculations)
    3. In a lot of use cases you will quickly discover that some data transformation or modelling can greatly enhance the power of your report. When using visualization only you are only using about 5% of what puts the Power in Power BI
  2. Visualization with Data transformation and Data Modelling
    1. This is the most common use case for Power BI and will give you about 80% of what Power BI has to offer without needing to write a single line of code – in other words this is not technical
    2. Although not technical; a good understanding about data and business intelligence principles is key to use Power BI effectively. The skills that are needed are outlined in the next section
  3. Advanced use cases; The advanced use case of Power BI adhere to the 80-20 principle: to effectively use the last 20% of Power BI will take you at least 80% of the time
    1. The most powerful element of this use case is the use of DAX. With DAX you can create very advanced calculations. Excel experts will feel some similarity to excel formulas; but the DAX language is conceptually more complex then any excel formula
    2. The second most powerful element is the Power Query or M language. When connecting sources and transforming data as done in use case 2, Power BI automatically generates this code. For advanced use cases you can alter or add to this code, for instance building a function that generates a date table from scratch
    3. The third element that is more advanced is setting up Power BI with additional Azure components for data transformation and / or storage. This scenario can increase the scale and performance of how Power BI can be used
    4. The last element is to build your own custom visualizations for use in Power BI. This requires good programming skills as the visuals need to be coded in the TypeScript
What skills do you need?

Below a short overview of which skills are essential to effectively use Power BI:

  1. Visualization only
    1. Advanced understanding of the business requirements and business value that the report should deliver
    2. Advanced understanding of the visualization options in Power BI. Partly because Power BI is evolving so fast; it is my experience that a lot of people who start off with Power BI know too little about the reporting possibilities. I would advise to follow at least the sections about visualization of the online courses
    3. Advanced understanding about data visualization concepts and best practices (e.g. visuals used, colors used)
    4. Basic understanding about Power BI data sources; how to connect to data
  2. Visualization with Data transformation and Data Modelling
    1. Advanced understanding about data structure and data models (read the first 100 pages or so of The Data Warehouse Toolkit by Ralph Kimball and Margy Ross)
      1. What is the grain of a table?
      2. Which fields are facts / measures and which are attributes?
      3. What is the cardinality between tables?
      4. How to design a star schema?
    2. Advanced conceptual understanding of the Extract, Transform and Load (ETL) process; For the data transform step; knowledge of different ways to change data:
      1. Filter
      2. Enrich with additional fields (nondestructive)
      3. Restructure ((un)pivot)
      4. Combine (join / merge, union / append)
      5. Aggregate
      6. Etc.
  3. Advanced use cases
    1. DAX: Good conceptual understanding of the DAX language; see the Blogs report for a lot of use cases and tips on how to use DAX. Apart from blogs there is alot of information on DAX on the Books, Courses and Tools pages.
    2. Language M: Advanced use cases in ‘M’ are best found in the Blogs report
    3. Azure: Whereas all the above topics are / can be completely Business generated (Self Service BI), when Azure components (databases, ETL tools, etc.) are added to the equation, the IT department needs to be involved (I would also involve them in the other use cases, but that’s beside the point)
    4. Knowledge of TypeScript which is a superset of JavaScript. See this page on the Power BI site how to get started building custom visuals

Getting Started - on this site

Regardless whether you are new to Power BI or an expert, if you are new to Power BI Central below leveling indicator will help you decide where to get started on this site.

Who you are: You are the Starter.

Why you are here: You just heard about something called “Power BI” and would like to know what all the fuss is about.

What you should do: Read these introductions on Power BI Desktop and Power BI Service. Go to the Videos page and watch Microsoft’s Power BI Channel. Go to the Tools page and download Power BI Desktop. Play around. Be thoroughly impressed and go to the Courses Page and take the edX course.

Who you are: You are (somewhat) skilled.

Why you are here: You were searching for one place that connects you to all available information on Power BI.

What you should do: Go to Books page and order a book. Go to the Forums page and ask a question. Go to the Blogs page and determine which Blogs you absolutely need to follow and wonder aloud how I made that Power BI Report. Join a group on the Groups page.

Who you are: Your native language is DAX. You are fluent in M. You think English is overrated.

Why you are here: It is your job to know everything. Hence you are here.

What you should do: Go to the Blogs page and think what you would have done differently in building that report. Search for very specific and complicated blog posts. Go to the Ideas page and check in weekly how the features you voted for are doing.

Share this page