3 Hacks for Better Dashboards

by | Dec 30, 2014 | General BI

Reading Time: 5 minutes

The purpose of business intelligence (BI) tools and dashboards is to provide a simple way to make sense of complex data. Unfortunately, when trying to visualize data, many dashboard designers add further complexity into the process. A complex, cluttered dashboard not only looks terrible, but it’s also of no business use if you can’t make sense of the data it’s showing you.

Here I’ll take a look at 3 dashboard hacks and how you can use them to produce a simple, clean dashboard that enables users to make quick business decisions. In other words, so you can let your BI tool do what it’s supposed to do.

Tweet: 3 hacks for better dashboards

Hack #1: Resist the temptation to be too “tricky”

Anytime I get a new toy, I feel exhilarated at the thought of trying out all the bells and whistles. Working on your first dashboard can produce the same feeling. Many BI software packages contain fancy graphical tools that allow the dashboard designer to change font color and size, add borders and backgrounds, or manipulate proximity and size of tables and charts. Unfortunately, if you don’t use a little self-restraint, it’s easy to get carried away in adding graphic embellishments. The result can be a dashboard design that is cluttered and confusing. This impairs anyone’s ability to read the dashboard and walk away with a clear understanding of the information disseminated.

Figure 1: This dashboard has way too many different font typefaces, treatments and colors, making it hard to read.

As a rule of thumb, I’d stick to a rule of “No more than 3.” That means don’t use any more than 3 of the following graphical parameters:

  • Font styles (font family, bold, italic, all caps, etc.)
  • Changes of shape (square cornered rectangle, round cornered rectangle, circle, etc.)
  • Colors (grays, blues, yellows, etc.)
  • Sizes (24 point title, 18 point subtitle, 9 point chart type, etc.)

Some people would cringe at the thought of using as many as 3 different font families. So remember: it’s OK to use only ONE font style … use one, two, even three styles. Just don’t use four or more. It will be too jarring to the eye.

Figure 2: This dashboard has only one typeface (Lucida Sans) used in only 3 different ways:
1. Variations of case: all caps or upper and lower case
2. Variations of size: chart size and heading size
3. Variations of color: black, blue or white

Hack #2: Don’t do the same drill-down over and over again

Once users get a hold of their dashboards, new questions arise and they start digging down into the data. This is a good thing! Business intelligence tools were designed to help you make sense of your data and answer the questions you have. However, over time you may find that you are doing the same digging (and digging, and digging) to get to the same data points.

If this is the case, it only makes sense to set your result to the top level of your dashboard so you have your answer right away without having to do the same drill-down over and over again.

Figure 3: Quickviews allow dashboard designers to facilitate the effect of “Diving” into data without changing screens.

Hack #3: Don’t settle for any old table or chart… make your data pop!

It can be difficult to fit a lot of data onto one dashboard, especially with the variety of media people can use to view their data. With an increasing number of users turning to mobile devices as their “go to” working device, the amount of dashboard real estate is shrinking.

One great way to deal with a smaller physical space is to summarize the information with a chart. However, there are times when individual numbers are just more important than summaries are. In these cases, tables may be your best option to display the data.

The type of chart you select is important too. Let’s take a look at one example…

Figures 4 and 5: Count of countries from which individuals have accessed the Dimensional Insight blog shown as both a pie chart and a line chart.

In the example above, the pie chart clearly shows that year 3 had more countries visiting our blog than did years 1 and 2. However, the use of the pie chart makes it difficult to determine whether year 1 or year 2 had hits from more countries than the other year. The line chart clearly shows that the spread of interest in our blog has consistently grown year over year and does not jump up and down in popularity.

In addition, pie charts should only be used to show percentages that add up to 100% or parts that add up to a whole. In this example, our countries represented in year 1 (2012) are most likely all also included in year 3 (2014), so the pieces actually add up to more than the total number of countries that have ever visited our blog.

For more information

Make sure the visualization type you are using is appropriate for disseminating the data the user needs to know. You may want to refer to “Selecting the Right Charts and Indicators for Visually Impactful Dashboards,” which is a great in-depth article on selecting the right chart type.

Creating dashboards can be time-consuming so you want to make sure your users actually use the end product. Use these 3 dashboard hacks and you’ll be on your way to a dashboard that is not only easy to understand, but also enjoyable to view.

Rose Curtis

You may also like