You can’t just throw a few Power BI visuals into a report and expect users to be able to analyse their data more effectively. In this post I try to highlight what I think makes an effective Power BI report.
Firstly, I recommend anyone starting to work with data visualizations and dashboards to read the book ‘Storytelling with Data’ by Cole Nussbaumer Knaflic – a simple introduction to the key principles.
Here’s a very simple report I created to highlight some of the considerations. You might want to view it full screen by clicking on the diagonal arrows at the bottom right hand side of the report.
The importance of a title
I remember when I drew my first graph at school, the teacher drilled into us the importance of every graph having a title describing what the graph was about. So I’ve chosen to include the name of my school, and the fact that the front page hosts all the most important statistics for my school. I tried to avoid buzzwords like ‘KPI’ or ‘metrics’.
Put the most important statistics first
In English, we read from left to right, horizontally and then down, forming a zigzag down the page. Hence, the data we place first is interpreted as being the most important. So I placed the total number of pupils on roll right at the top left corner, followed by boys and girl counts and a year group count. As we move down the page, the graphs are ‘less’ important.
Use colour sparingly
In this example, I’m just using a single colour. The idea is that I will only use different colours to highlight an important difference between measures. On this page there is no need to use any colours other than green and black (for the text).
Tell the viewer what they are seeing
I have a text block in the middle of the report to guide the viewer. I keep the text short, and to the point, but it hopefully puts the viewer in control of the report and helps them to interpret the data correctly.
Be careful with pie charts
I could have used pie charts for a couple of visualisations (pupil premium and SEN for example) but chose not to. Pie charts can be poor ways of showing the very small slices of data. Pie charts do have their place, but often bar charts are better.
Provide extra help when using slicers
Many school staff have problems using slicers for the first time. I’ve clearly identified all my slicers (and called them ‘filters’ because many school staff find filters more intuitive) by putting them in a box at the right hand side of the screen. I used text to describe how slicers can be used to make multiple selections. And I’ve provided a button to help users clear the all the slicers.
Do you really need a Y axis?
In the comments below, DAX Fan asked ‘ If you are using data labels, why use the Y axis?’. That’s a very good point. In the first version of these visualizations I included the Y axis and the data labels. Now I just show the data labels. The visualizations tell me exactly how many boys and girls are in each class – so I don’t need a set of horizontal lines to tell me the same information.
Finally, indicate how old the data is
School data needs to be up to date – fresh – and timely. If you look in the top right hand corner of the report I’ve included a ‘last updated’ data and time. Users of this report will know exactly when the report was last updated.
Contact me if you want me to build a report like this for your school. Meanwhile, feel free to add your suggestions to the comments below.
If you are using data labels, why use the Y axis? When there aren’t too many data points, it’s preferable to use data labels and remove the Y axis (or X axis in case of horizontal bar charts). Try it — your report above will look much better.
Good tip! And a good point too – I can see that it just replicates data and increases screen clutter.