Power BI design tips

5 Tips for #PowerBI

After a couple months of fun with Power BI, I’ve picked up a few little tricks along the way that have helped me to be able to create some pretty cool data visualizations and dashboard reports. Here are five Power BI tips and tricks that you may find useful as you begin creating dashboards for your organization.

New to Power BI? Start here to get acquainted!

1) Use a pie chart or donut chart as a KPI

One of the ways we can create a KPI visualization is to use a pie or donut chart visualization, which you can see here.

image

In the chart above, I create a KPI to quickly display which tight ends score more than, less than or equal to the average number of touchdowns all tight ends scored last year allowing me to quickly identify tight ends that score more TDs than average.

Here is my calculated column to create the KPI value for you to have as an example:

TD KPI = if(int('TE Stats'[TD])>INT('TE Stats'[Avg TE TDs]),"1",IF(int('TE Stats'[TD])<int('TE Stats'[Avg TE TDs]),"-1","0"))

 

Then configure a pie chart as follows.

First, I place my KPI calculated column as the Legend and as the Values.

Power BI KPI pie chart

Then I hid all the labels and configured the colors to display red (-1), yellow(0) or green(1) depending on the value of the KPI.

Power BI KPI pie chart

Now when I use a slicer to select a player, my pie chart acts as a stoplight KPI. Cool!

Power BI KPI

2) Use a chart as a slicer

I’ve previously blogged this tip before, but this one is too nifty to not share again, in my humble opinion. One of the advantages to using a visualization like a funnel chart as a slicer is you gain the ability to single-select a filter, which is something the current slicer lacks. Check out this post to learn more about leveraging Power BI’s natural cross filtering to create some pretty cool slicers.

image

3) Create a spark line with a line chart

A nice way to create a small trend line, also called a spark line, is to use a line chart visualization which you can see below.

Power BI sparkline line chart

This trick is pretty easy. Just create a normal line chart visualization, hide all the labels and shrink the chart down to the desired size.

4) Use a scatter graph and matrix to create a calendar chart with day labels

Last week you may have seen my blog post on how to use a scatter graph to create a calendar chart. One of the ways you can improve the calendar scatter graph is to create the visualization along side a matrix visualization, as seen below.

image

You can use the matrix to display the names of the week below each day in the calendar and then also optionally display the totals by day.

5) Right align the y-axis on a bar chart to prevent the labels from hiding

The bar chart is a great visualization type to use in your Power BI dashboard because its so easy to differentiate the differences between the categories. But one of the problems with the visualization in Power BI is that sometimes its hard to see the categories on the y-axis if the chart is too small. See the image below to see an example the issue I’m talking about.

image

One way you can work around this is to right-align the y-axis. This will cause the full value of the y-axis categories to always be displayed in all their glory albeit on the right side.

image

You just have to live with the category labels on the right side of the bar chart.

Resources

Need more Power BI tips? Check out these tips:

Here’s three Power BI best practices to follow.

Here are the new visualization types in Power BI.

Converting Power Pivot models to Power BI is now a thing!

16 thoughts on “5 Tips for #PowerBI”

  1. Thank you for your post! Do you know of a way we can dynamically add a date to a the string text of a Report Title? Is it possible to blend a string text and DAX in a Text Box? For Example: “Production Report 1/1/2015 thru 2/1/2015” which would have 1/1/2015 and 2/1/2015 pulled from our PowerPivot DataModel as a Source?

    Thank you in advance for your thoughts on the above!

    1. Hi, Mark. Thanks! The way I’ve seen this accomplished is to create a measure in the model that displays the date. So your measure expression would something like this: Begin Date = format(MIN([Date]),”MM/dd/yy”) and you could also create one for your End Date. Then display the measures in a Card viz like seen here: //sqldusty.com/wp-content/uploads/2015/09/capture-e1441378811822.jpg
      I hope this helps. 🙂

  2. Dustin, nice touch with the pie charts…how did did you get the value to display in the middle of the pie chart as opposed to a data label outside the chart

  3. very interesting, it’s not very clear how you managed to obtain the set of players, each with an individual pie chart and his name. If you use a slicer, wouldn’t you end up only with the selected player? Thanks

    1. It’s been a while since I looked at this report, but in this case I believe I used a visual level filter on these visualizations to filter down to a single player.

Comments are closed.