September 8th
February 22nd
Power BI is, without doubt, one of the most advanced and useful business intelligence tools ever built. But when starting with Power BI, many people feel equal parts excited and overwhelmed: excited because it opens the doors to so many opportunities, overwhelmed because Power BI can be, well, a lot (at first).
In Webinar 47 (pfew, it's been a while!), Tilen and Mark untangled the seemingly mysterious ways of Power BI to build actionable reports. Here's a quick rundown of what they discussed (you can also watch the recording using the button below, as always).
Building reports in Power BI may seem straightforward at first—you import your data and start creating visualizations. However, this approach often leads to complications, especially if you’re working with flat tables or multiple disconnected tables. To truly harness the power of Power BI and simplify reporting for everyone in your organization, it’s essential to follow a structured process.
Here are two crucial steps to set you on the right track, no matter what you decide to do with your data next:
The first step is data preparation. This means cleaning and transforming your data so it’s ready for analysis. Skipping this step can lead to confusion and inefficiency later on. Although data prep might initially seem complex, it’s actually quite manageable and necessary to ensure a smooth start.
Creating a proper data model, often by using the star schema, is critical. This approach works for about 99% of semantic models and simplifies subsequent processes like writing DAX formulae. A good data model provides a solid foundation for your reports, making both building and interpreting them much easier.
Once these steps are complete, the focus can shift to visualizing the data. By using actionable elements such as variance analysis, KPIs, and target tracking, you can create reports that present insights clearly. The goal is to design reports that users can understand without needing an expert to interpret for them.
Organizing your data into a star schema ensures optimal performance and scalability. In the webinar, Mark suggested you follow these steps:
If you're starting with a single consolidated table that contains all necessary data, follow these steps:
Importing the data in Power BI is a very easy, two-click process:
Although you can, theoretically, not transform your data and use it as you have it, it is recommended you do clean and transform it first, so you can start working with a proper dataset from the get-go.
Here's how Mark did this during the webinar:
During the webinar, Mark went on and started implementing a Star Schema data model. This is necessary because it ensures a good performance of your Power BI reports, even as you scale your solutions.
The first step towards that is by referencing the clean data table and working on it by establishing the time intelligence (this is the first part of doing this, we will come back to it further down the process to finalize it).
To do this, you have to start by identifying or creating date columns. In the webinar, Mark demonstrates this by creating a date column from Columns built for Year and Month indicators, by using the merge function and building a column from examples functions.
To avoid repeating columns in text format, you then have to separate the dimensions from the table (as a general rule, the text format is not as efficient as the integer/ number format, particularly when it comes to data model performance).
To do this, follow these steps:
To bring the dimensions back to the fact table (without the text columns), merge back the dimension table to the fact table.
Remember to also remove dimensional text columns (you can do this by using the "Remove other columns” function).
This is one of the last steps of the process, as the result of doing this will be your Star Schema model. Before moving on, however, you will have to...
To do this, Mark suggests creating a calendar table using DAX code (we have a full article on how to create a calendar table using DAX, here). This will allow you to leverage the Power BI time intelligence to calculate periods like PY, YTD, MTD, and so on.
Want to see this in action?
Once you have your Star Schema in order, you can proceed to build advanced visualizations on your Power BI dashboard. Below are some key bits of information you need to know about this, as presented in the webinar.
There are, of course, a lot of data visualization best practices to keep in mind. However, for the purpose of building this actionable report, we suggest you to:
As a bonus point, the Zebra BI cards erase the need for additional slicers as these can act as slicers as well.
With Zebra BI visuals, you don't need to worry about how to design and how to ensure proper formatting of each element. Visuals can take care of it and save you time.
When presenting financial data, the table format is recommended for a number of reasons:
Here are a few things you can implement in a table format to elevate your financial data reports:
You can turn ERP exports or similar datasets to Power BI reports as part of your POC.
To do this, you should follow the Clean and Transformed Steps mentioned earlier, during Stage 1. Additionally, you can also do what Tilen showed webinar attendants, specifically for financial use cases:
To better visualize Profit & Loss statements, follow these best practices suggested in the webinar:
Essentially, turning your Power BI reports into actionable dashboards boils down to five steps:
Start by cleaning and transforming your data. In the webinar, Mark and Tilen have shown you how to follow some basic steps in Power Query that are appliable to most common use cases.
It might look overwhelming, but it isn’t as complex as many think.
The key here lies in implementing a proper star schema data model. Setting up a proper data model is important for enabling future scaling of reports from POCs projects to business reporting solutions used across your organization.
Adopting business communication best practices (such as the IBCS) will help you make sure your reports are understandable and standardized. This also means consistency and clarity for report end users.
Having the right set of tools or software to implement any kind of business is an enabler for ROI on such projects or transformations.
This way, the report developer doesn't have to care about the IBCS rules. You just need to prepare the data and the software makes sure the visuals created from the data are already aligned with IBCS standards. This is a win-win scenario for report developers and end users.
In the Power BI tutorial we have shown, a lot of emphasis was on variances. The reason behind this is because they give context to your figures: is your performance good or bad? And how good or bad? Being able to answer this kind of questions from a report can be a gold mine of insights for your business.
Using advanced visualizations in Power BI helps you bring actionable reporting elements to life. Examples include:
Annotations can make all of them even more insightful as they give you the opportunity to explain the “why” behind the numbers, which will make your reports very actionable.
As your report (your POCs) is now build on very strong foundations, you can start scaling your solution. You can:
Organization-wide adoption will create a data-driven culture—and that is precisely what you want.
Take these with you, put them into practice, and deliver reports that stimulate actions and create value for your business. If there’s no action, then all reports, all data warehouses, are just costs.
Data becomes valuable only when a decision is made and an action is taken, and the steps described in this webinar will help you do that in a clearer, more straight-forward, and accurate way.
