DataConsulting’s Data Insight Analyst Jonathan Jedrczak shares with us his “Power Thoughts” in the first of a series of blogs:-
Top Tips for Dashboard Design
Anyone can put some visuals on a page and call it a dashboard. However, the challenge is making that dashboard look stunning, whilst still being neat and professional. In this first article for “Power Thoughts”, I am going to discuss this topic by showing examples of a dashboard during 3 stages of its development, along with a discussion of the ideas behind the improvements.
V1 – Starting Dashboard
V2 – Improved Dashboard
V3 – Finished Dashboard
1 – Layout/Grouping
When designing dashboards, it can be useful to imagine splitting the page into rectangular sections. This way, it guides the user on which visuals fit together and generally appears neater. In V2 – Improved Dashboard (V2), the left-hand rectangle is a collection of stats and slicers. The middle is information relating to geographic location and finally the right focuses on demographic. Another point on layout is the importance of space. It’s important to use the screen efficiently to communicate information, but not include so many visuals it becomes muddled and confusing. By grouping the visuals using the technique described above, it becomes easier to judge when this balance has been reached.
2 – Colour
It’s advised to choose a colour palette up front. To help with this, you can use “Themes” in Power BI, which are prechosen groups of colours that you can restrict yourself to when designing your reports. Having chosen your theme, you can then think about implementing it. Ideally, there should be good contrast between the background and visuals/text to help them stand out. Personally, some of my favourite colour palettes revolve around dark backgrounds and light text. In particular, I really like a dark blue background with white text/visual backgrounds. Finally, In V1 item 2 there was an unnecessary use of additional colours. These didn’t add any meaning and therefore were replaced with just a standard blue, as seen in V2 item 2.
3 – Visual Choices
When trying to create an impressive report, an easy trap to fall into is using every fancy visual available. In some cases, this can be very effective but as a rule of thumb simpler is often more effective. In this case, using a waterfall chart to display the range of age groups is fine and doesn’t look that bad, but I’d prefer to see the same information in a column chart. This is shown in V1/V2 item 1. The “Population by Country” visual has also been improved. In V1 item 3 it’s hard to read the axis, so this visual has been changed to a bar chart. In addition, only the top 6 countries are shown to help highlight the key information.
4 – Consistency/Finishing touches
In my opinion, for a dashboard to look professional and slick, every little detail needs to be consistent. In V2 item 6, the visual header has a different font to the rest of the page. This is an obvious example but illustrates how little differences can have a big impact on the overall quality of the dashboard. This rule applies to visual size, text size, text font and anything else you can think of. Lastly, the finishing touches are all to do with alignment. In V2 visuals 4/5, there are several elements where they don’t quite align with each other. I strongly believe it’s worth the time of going through every visual and ensuring it’s aligned with everything else on the page. This can involve resizing, or even potentially having to reorganise the page entirely, but in my experience the end effect is worth it.
Summary
As a final point, none of these are hard fast rules, and as with every rule there’s always exceptions. These are just general guidance I find useful to bear in mind when developing dashboards, and I hope they may prove useful to you too.
