Power BI is a useful tool in a wide range of fields and industries because its dashboards and reporting functionalities can be tailored and implemented to reflect your company’s exact requirements. With this being the case, there are some really unique and useful dashboards out there with some of the best being found in the Power BI Data Stories Gallery.
We spoke to our Director, Stuart Box, who has put together a guide of useful elements to consider when creating dashboards:
Relevancy is of the first importance.
It doesn’t matter how pretty, or clever, or cool a report is, if it isn’t relevant to me, it isn’t important. Most of the gallery examples above fall into this category. Their primary purpose is to entertain and inform, they don’t really tell me anything important.
Now a report analysing my last years income, and projecting next years, is important and relevant. I’ll pay that report much more attention. Relevancy is of the first importance. If the numbers I want are buried in a whole load of irrelevant data, then that report is more difficult to understand and hence, less useful.
Users shouldn’t need to think.
The report needs to be clear to be useful. There’s a wide range of visualisations in Power BI and it’s easy to get carried away and show the data in multiple ways, without making it easy to understand.
‘Don’t make me think’ was a maxim coined by Steve Krug back in the early days of website design, but it still hold true for reports. The report should be clear in how it’s used, how does the user slice and dice the information presented. It should be obvious, clear and easy; the users shouldn’t need to ‘think’.
Thoughtful use of colour to highlight areas of interest, outliers or issues is a good case in point. Some users just throw as many colours as they can at a report to make it look ‘bright’ or ‘cool’, but actually a more restrained pallet may make for greater clarity.
It seems obvious that the data on a report should be accurate, but you’d be surprised how many reports I see that aren’t. Pie Chart Percentages that don’t add up to 100, Grand Totals that differ from the sum of the data on display, Subtotals that don’t. These are all common problems, see our post on DAX: Is the Total Row in a table giving you grief?
DAX is a very powerful language, but it can be misunderstood and return incorrect values which an unsuspecting report writer simply believes. Reports should always be cross checked for accuracy. This may not be an easy job if the underlying data contains millions of rows, but it is vital.
In extreme cases reports should be auditable, explaining what calculations they are performing, how they act on the data, and providing in-built cross checks to prove accuracy.
Clarity and accuracy provide a basic beauty of their own.
This comes last, for possibly obvious reasons. I can forgive an ugly report if it is relevant, clear and accurate. Indeed, clarity and accuracy provide a basic beauty of their own. I don’t necessarily need a pretty background, or flashy animations if the reports gives me the basic information I want in a clear and accurate way.
That said, if I’m going to interact with a report regularly, say a weekly Sales report or a daily Expenditure report, then it’s good if that report looks as pretty as it can. Also if I have to sell, inform or persuade with a report then good use of graphics and colour will help.
Of the many authors writing books on Dashboard and Report design, Steven Few and Alberto Cairo have influenced my own design attempts along with Guy Kawasaki and Garr Reynolds. If you get one book each from these authors you’ll have a great background in the design of good looking reports.
Learn how to create powerful dashboards
Of course to understand how to implement this design with a tool like Power BI, you’ll need good training. To find out more information about our Power BI training courses, call us on 0800 0199 746.