Moving Dashboards out of the 90’s PowerPoint Phase

September 12, 2013 Jon Hazell

Remember PowerPoint in the 90’s? The near seizure invoking presentations that people put together with slides and text flying in or dissolving from every angle? Huge 3D fonts with massive gradients and colors reminiscent of that time you “accidentally” ate the wrong thing at Woodstock?

A far cry from the less is more approach that is taken today. Now presentations are minimalistic, clean, and, surprise-surprise, easier to understand. The message isn’t lost in the delivery. How you are presenting isn’t taking away from what you are presenting.

Dashboards, for a while now, have been stuck in that 90’s PowerPoint phase, and almost everyone has been guilty of aiding this phase. People wanted that wow factor of 3D effects, sweeping gradients, or images in the background of their dashboard. Pie charts are used because people are used to seeing them, and for some reason people really like circles (think bubble charts). But with putting in these bells and whistles, dashboards are damaging the one thing they are supposed to be the best at doing – simplifying how we consume data/information. How we are delivering the message has been detracting from the actual message itself.

We can see the light at the end of the tunnel, and everyday it’s becoming brighter.
As people become more and more aware of data visualization best practices, we are starting to see a shift away from this 90’s PowerPoint phase. For example: Dundas doesn’t offer 3D effects in Dundas Dashboard. Why? Because when you make something 3D it makes it hard for the viewer to understand what’s happening. Take a look at the example below:

There are a lot of things wrong here:

1) The area becomes distorted on the pie chart when 3D. Pieces that should be the same size appear bigger or smaller depending on positioning.

2) The positions appear to be arbitrary when they should be in ascending or descending order according to their size.

3) A pie chart generally should not be used when more than 5 (or according to some people 4) values are being displayed. A bar or column chart can be more affective then.

With a dashboard, you want to display all the relevant information you need on one page. Ideally you shouldn’t have to scroll down in order to view more information, this is known as displaying above the fold. You also should be able to understand what it is that you’re viewing within a matter of seconds (not minutes). This means effective use of labels, data visualization best practices of showing and not distorting the data (like the case with the sample above or the lie factor discussed in Using Dashboards for Good or Evil), and color.

When presenting data, it is nice for it to look good, especially when you’ve paid good money for a dashboard product. But looking good, and looking gaudy, are two different things. Gradients should be very minimal, if used at all. Designers should be aware of when they are using red and green hues because of those who are color blind.

So proper color, no 3D, no lie factor, and proper labels are all good right? But how in the world do you get all the info crammed into a page? Well, there are couple of great visualizations for this: Sparklines and Bullet Graphs.

Data Visualization guru, Edward Tufte, describes a Sparkline as “data-intense, design-simple, word-sized graphics”. You generally want to use a sparkline when displaying data over time, and they are often used in conjunction with text.

Bullet Graphs are a variation of a bar chart developed by another dataviz guru, Stephen Few. Few developed bullet graphs to overcome the fundamental issues of gauges and meters, as they typically don’t display enough information, require a lot of space, and are cluttered with useless and distracting decoration. The bullet graph features a single, primary measure (e.g. current month-to-date revenue or output), compares that measure to one or more other measures to enrich its meaning (e.g. a target or threshold), and displays it in the context of qualitative ranges of performance, such as poor, satisfactory, and good.

The image below contains both sparklines (on the left) and bullet graphs (on the right).

It would be easy to just say ‘keep it simple’, but that would be a misnomer. Dashboards can be quite complex and sometimes in order to better understand the data being viewed you need to provide more information to add context (annotations, hover overs). What I like to say is keep it clean. Display information in a way that is easiest for the viewer to understand. Don’t make a dashboard more complex than it needs to be. Keep space in mind in order to maximize it (your company logo may not need to take up a third of the screen). Keep animation and gradients to a minimum (yes they add the wow, but do they truly add value). Finally don’t muddy up your data by distorting it.

Previous Blog
The Perils of Big Data
The Perils of Big Data

Businesses are currently told that their data is one of the most important commodities, which means everyon...

Next Blog
Why Data Visualization Matters
Why Data Visualization Matters

Visualization lets us see things that would rather go unnoticed. In visualizations, we can see patterns tha...


Want more BI tips, trends, and insights? Subscribe to our newsletter!

First Name
Last Name
Company Name
I consent to receive commercial electronic messages from Dundas regarding products, services, updates and other information about Dundas. You can withdraw your consent at any time.
I Agree
Thank you!
Error - something went wrong!