4 Tips for improving your KPI dashboards
Designing a KPI dashboard can be challenging to say the least. Choosing how best to communicate many important metrics, without overwhelming your end users, can be difficult. Here are some tips and tricks to help keep your dashboards engaging and looking sharp:
1. Sometimes a number will do
With the shift from Excel to BI packages, it can be easy to fall into the trap of trying to turn everything into a visualisation. But sometimes just a number will do the trick. Particularly for headline figures like total sales, profit and units sold. It’s best to put these leading KPIs at the top of your dashboard as they’re probably the best indicators of the overall health of your business. To highlight these numbers even further, put them in bigger and bolder text and add a boarder around them to draw your user’s attention.
2. Limit excessive detail
Does it really make a difference that your sales for last quarter were €200,974.65 as opposed to €201K? Displaying excessive detail on your dashboards can cause your visualisations to look busier than they need to be and can overwhelm your end users. Remember, you’re designing for top line KPIs, and additional detail can be added into a drilled down version of the report. If you’re working with an interactive tool you can add additional detail into a text object, which pops up when hovering the cursor over the view.
3. Pie charts aren’t (always) evil
There’s a preconceived notion among the data visualisation community that pie charts should be avoided at all costs. This is not because pie charts are inherently bad, but because they are often used incorrectly. The human eye isn’t great at deciphering small differences in angularity. This is why more often than not it is better to opt for a bar chart, where small differences in length are easy to identify. Similarly, when using a pie chart, you should always be comparing a part of a whole. For example, you should never display your top 5 sellers and their percentage of sales in a pie chart. This would be misleading, as it could be interpreted that these top 5 sellers represented the total sales for the business. Pie charts are particularly effective when comparing 2 measures as a percentage of a whole. Think online sales vs store sales. As a general rule of thumb, don’t have more than 4 measures in your pie chart, otherwise the differences between the slices become difficult to decipher.
4. Use colour sparingly
Data visualisation tools today make it easy for us to add colour to every element on our dashboards. It’s easy to get over zealous and add colour without properly thinking it through. When adding colour, ask yourself; Is it really necessary? What is it adding to the visualisation? Colour should be used sparingly and should only be added if it has a purpose. Keep your colour palate simple and stick to 3-4 colours max. Remember grey goes with pretty much anything, and softer colours are easier on the eye. Also keep in mind the media of delivery for your dashboard. What screen resolution do your end users have? Will your dashboard be printed in black and white? These are important considerations to test for during the development phase.
