Top Business Intelligence dashboard design best practices (Part One)

23 Min Read

Don’t you hate it when people gripe, whine and bemoan but don’t offer a solution to the source of angst?

Recently, you might have mistaken – or possibly even rightfully labeled – me as one of those chronic-no-help-complainers. Like Prima Donna singer-songwriter Elton Hercules John.

Don’t you hate it when people gripe, whine and bemoan but don’t offer a solution to the source of angst?

Recently, you might have mistaken – or possibly even rightfully labeled – me as one of those chronic-no-help-complainers. Like Prima Donna singer-songwriter Elton Hercules John.

I’ve discussed, with notable overtones of dismay, the persistent and continuing problem of poor Business Intelligence (BI) dashboard design and delivery, and it’s ability to work against the very purpose of BI itself – to attain accurate fact-based insight into the operational (or strategic) performance of an organization and facilitate action, derived from that knowledge, to improve processes and achieve competitive advantage.

But it’s high time that I elucidated this potentially disastrous situation with something useful. Some suggestions.

A working definition

To begin with we need, at the very least, a working definition of a BI dashboard to ensure that any best practice comments are contextualized.

Searchcio.techtarget.com defines a dashboard as “a user interface that, somewhat resembling an automobile’s dashboard, organizes and presents information in a way that is easy to read. However, a computer dashboard is more likely to be interactive than an automobile dashboard (unless it is also computer-based). To some extent, most graphical user interfaces (GUIs) resemble a dashboard. However, some product developers consciously employ this metaphor (and sometimes the term) so that the user instantly recognizes the similarity.”

Another of TechTarget’s enterprise technology resources, searchbusinessanalytics.techtarget.com, offers a more specific description: “A business intelligence dashboard is a data visualization tool that displays the current status of metrics and key performance indicators (KPIs) for an enterprise. Dashboards consolidate and arrange numbers, metrics and sometimes performance scorecards on a single screen. They may be tailored for a specific role and display metrics targeted for a single point of view or department. The essential features of a BI dashboard product include a customizable interface and the ability to pull real-time data from multiple sources.”

And, to avoid additional confusion (or if you weighed into the “Difference Between Balanced Score Card and Dashboards” debate in the Business Intelligence Professionals LinkedIn Group), the SearchBusinessAnalytics.com definition also differentiates dashboards from performance scorecards:

“The business intelligence dashboard is often confused with the performance scorecard. The main difference between the two, traditionally, is that a business intelligence dashboard, like the dashboard of a car, indicates the status at a specific point in time. A scorecard, on the other hand, displays progress over time towards specific goals. Dashboard and scorecard designs are increasingly converging. For example, some commercial dashboard products also include the ability to track progress towards a goal. A product combining elements of both dashboards and scorecards is sometimes referred to as a scoreboard.”

However, the most succinct and workable definition of a BI dashboard still belongs to data visualization expert, Stephen Few. Few distills the concept of a dashboard in a single sentence: “A dashboard is a visual display of the most important information needed to achieve one or more objectives; consolidated and arranged on a single screen so the information can be monitored at a glance” – Stephen Few in “Dashboard Confusion”, Intelligent Enterprise magazine, 2004 March 20.

Now, before we launch into some of the top best practices, it’s worth remembering that a BI dashboard is a delivery mechanism for sharing information with a designated user or group. It is not pursuant to one particular type of technology or information category – despite the fact that most dashboards display metrics relevant to specific KPIs.

Best practices for dashboard design

1. Make communication the number one priority

It’s not about how pretty they are, it’s about their ability to quickly and clearly communicate the right information in the best way possible to enable better, faster fact-based decision-making. In short: Pretty is not the objective. Communication is the objective.


2. Modern dashboards must be Web-based

Contrary to more traditional definitions and best practice examples of BI dashboards, delivering dashboards via any other means except the Web is more than old school – it’s obsolete and archaic. How else are dashboards meant to deliver close to real-time information to empower you to grasp an opportunity or avoid a crisis? How else can you share insights and spark instantaneous conversations with other involved decision-makers and prompt fast action? Everything else in your business is online; from your CRM to your HR system – it’s just quicker, smarter and more convenient. So, what do you do if your dashboards aren’t online? Get with the times!

 

3. Fit it all on a single screen, without sacrificing meaning, wherever possible

Whilst there will be some circumstances that dictate otherwise, based on the particularities of the information displayed, all relevant reports and information should be displayed on a single screen where possible. Doing so, providing that clarity has not been sacrificed, allows for faster data inspection and understanding. Cognitive research – studies examining how the human brain processes information – has demonstrated that the ability to visualize all interrelated information sources together, empowers people to more easily and clearly grasp significance and overall meaning of an information set with greater accuracy.

 

It also allows for easier and faster comparisons between different chart types, as well as the identification of trends and relationships within the overall data set, leading to deeper insights that might otherwise not have been attained.

A dashboard that’s longer than the length of a screen, and requires scrolling, is less effective. This is because the brain is only capable of storing a small amount of information in short-term memory, as indicated in the results of a study by Christopher A Sanchez, published in The journal of Human Factors and Ergonomics Society, October 2009 (vol. 51, no. 5 730-738). The study – “To Scroll or Not to Scroll: Scrolling, Working Memory Capacity, and Comprehending Complex Texts” – aimed to assess the impact on peoples’ ability to process complex information via scrolling Web-based interfaces.

Sanchez found that: “Results from both studies indicated that a scrolling format reduced understanding of complex topics from Web pages, especially for readers who were lower in working memory capacity.

“These findings show that the way text is presented can interact with learner abilities to affect learning outcomes.

“These results have implications for both educational technology and human interfaces that present information using displays that can vary in size and construction.”


4. Start with the big picture, drill to detail

Information should be displayed in a series of high-level summary reports that quickly provide an overview of what is happening in an organizational environment, which allow users to drill down into subsequent detail to diagnose why something has occurred and reveal its genesis.

 

5. Ensure natural alignment and logical order of charts

A 2003 study by Robert Pearson and Paul van Schaik, published in the International Journal of Human-Computer Studies, found that Web page layout was integral to people’s ability to quickly and accurately comprehend information. The research, entitled “The effect of spatial layout of and link colour in web pages on performance in a visual search task and an interactive search task“, found evidence that supported the notion that experienced Internet users have formed “automatic attention responses” and visual search patterns based on commonly accept norms of Web page or Web interface design. Therefore, as nowadays almost any recipient of a BI dashboard would be classified as an “experienced Internet user”, the position of graphically significant information within a dashboard should conform to the expected norms of Web page layout to assist information comprehension.

For example, any menus should be left adjusted where possible, additional tabs should be positioned across the top of the dashboard display, and if scrolling is necessary, then users should scroll up and down, rather than across the screen.

In addition, charts within the dashboard should be ordered in a way that enables the fastest, easiest and most logical consumption of information. For example, positioning a series of five reports on a dashboard, where the significance of the first cannot be fully understood until the user has read the fourth, equates to poor design as it slow information comprehension.


6. Customization is the key

The reports and associated visualizations should be customized to suit the needs and specific demands of the recipient (individual or group), relevant to the functional role of that entity to the issuing organization, to support appropriate and timely decision-making and drive action. What would be the point of letting your marketing team know how your HR initiatives are performing?

 

7. Exceptions and alerts

Dashboards can be an invaluable mechanism for monitoring business performance across an enormous range of measures. However, dashboards by their very nature should be designed to convey at-a-glance-insight. A good dashboard communicates the right information in an easily understandable and actionable way – the whole point is that they’re designed to increase workplace efficiencies by removing the need for people to manually monitor relevant business functions and developments. So, to support this aim, users must be able to schedule a system of exception and alert notifications based on the reports within their personalized dashboard.

Exception and alert reporting enables users to set-up triggers based on predefined measures.

For example, a sales manager could set-up an automatic trigger action that stipulates that an event notification will automatically be sent via email as soon as quarterly sales fall below 80 percent of the sales attained for the same time last year.

8. Effectively highlight the most important information

Best practice dashboardsshould be designed so as to draw the users attention to the most pertinent pieces of information within a dashboard. To achieve this, you should avoid:

  • Cluttering the dashboard with visually gratuitous chart types that result in an assault on the eyes, and therefore failure to highlight / elevate a single point(s) of importance above other data displayed
  • Selecting a vendor whose dashboard interface is visually gratuitous and therefore distracts / detracts from the data being displayed

 

9. Use color appropriately and sparingly to achieve maximum contrast

The usability and visual impact of a dashboard are not necessarily compatible ventures. But, the careful use and selection of color when designing a dashboard of reports can improve the readability of charts, drawing the users’ attention to important changes, trends and measures.

Two prominent academic studies on the use of color and visualizations within Web pages – The effect of text and background colour on visual search of Web pages and Determining Users’ Perception of Web Page Visual Complexity and Aesthetic Characteristics – indicate that higher contrasts between colors leads to faster information searching capabilities and information absorption. In short, make your data points stand out from one another, and from chart and dashboard backgrounds and interface.

However, if all colors chosen to represent different metrics or values within a chart are eye-catching, no single point will standout above the others. Likewise, the presence of too many greatly distinct colors causes visual clutter and obstructs information interpretation. The end result is a visualization that is difficult to comprehend. Gratuitous use of bright colors undermines their intended value and purpose, resulting in a meaningless barrage.

Meaning must be derived from dashboards quickly. If the colors chosen make charts or graphs appear overly garish or complex, interpretation is hindered and usage rates will decline.

Colors should also be selected based on:

  • A clear understanding of their inherent or commonly accepted symbolic or metaphoric meaning (red = bad, etc)
  • The interrelationship between individual items on a chart and how that chart relates to all others on a dashboard (For example, if data relating to second quarter sales is displayed in purple in one chart, all other charts that display data relating to second quarter sales result should also be displayed in purple)
  • The goal of achieving maximum contrast, whilst avoiding color clashes and over colorization. For example, a chart using fluorescent pink, yellow and green achieves contrast, but clashes because it’s visually overwhelming. On the other hand, deep red, blue and grey contrast effectively without delivering a visual assault that hampers information comprehension.

10. Select the best, not the best looking, visualizations

Intuitive, immediately understandable and context appropriate visualizations – that convey the intended meaning and communicate that message clearly without occupying unnecessary or disproportional screen space (relative to the contextual importance of that KPI and its associated message) – are best.

Additionally, never include potentially distracting and superfluous imagery where a simpler alternative will communicate the significance of the data with more or equal effectiveness.

Remember that popular and pretty guy or gal from school – the one who never knew the answers, but was stunningly gorgeous? Bet you never asked them to explain that complex math assignment did you? You knew that they weren’t the best vessel to communicate that information. And, even if you did, chances are you were completely and utterly distracted by their appearance and failed to absorb any of the actual information they were trying to deliver to you.

The point? The data; not the dashboard, should always be made the center of attention. Never use flashy visuals and chart types when simple alternatives are capable of conveying the same message – does the third dimension on that pie chart really add to its meaning?

10.1 Reduce the data to ink ratio
Follow the advice of Edward Tufte in his renowned The Visual Display of Quantitative Information: Remove anything that isn’t absolutely central to the interpretation of the data. Only display objects that are vital to the accurate interpretation and contextual understanding of the underlying data – avoid all design aspects that are unconnected to the task of analytic communication.

“Perfection is achieved, not when there is nothing left to add, but when there is nothing left to remove”

– Antoine de Saint-Exupery

10.2 Select the right visualization for the data and the context
Selecting the most context appropriate visualization for a particularly metric or measure requires the judicious application of a little common sense. For example, if you’re attempting to monitor or track the change in something over time, a line graph will almost always work best. Likewise, if tracking several metrics of similar proportions – a potential example might be new leads generated for the current year by marketing category (Google Ads, LinkedIn, print media, banner advertising, etc) – using a column chart or bar graph would be an effective way to visualize the minor differences in performance between each marketing channel. Conversely, a pie chart would deliver a poor user experience as, at first glance, all the portions would seem equal.

10.3 Learn to love horizontal chart types
The visual orientation of the dashboard and its associated chart types is critical. Horizontal bar graphs and sparklines are simple-looking types of data visualization, but are highly effective for communicating information quickly and facilitating fast information absorption. The reason for this is that the human eye and mind are trained to read across a page from an early age.

 

11. Provide context

Without context, data visualizations have very limited usefulness. For example, a dashboard tracking supply-chain effectiveness might contain a report that displays the number of timely stock deliveries – as a percentage of total deliveries made – for the current month or quarter. On its own, it’s of very little value. Is the result communicated good or bad? How does it compare to the previous months; and is it trending up or down? Does the percentage displayed meet or fail any predefined service standards? Without additional contextual information to help answer such questions, it’s impossible for a user to understand the true meaning of the result, what action it requires, or whether it demands any action at all.

The addition of contextual and interpretive information, such as:

  • Labels;
  • Scales;
  • Headings;
  • Comparisons between other relevant or associated measures (such as the previous month, quarter or year results); and
  • Chart tool tips (the ability to hover over a visualization on a dashboard and be presented with a pop-up box containing additional textual information about that visualization),

Enables users to perform a meaningful interpretation of the information presented, and understand its implications and impact on organizational operation or strategy, empowering them to take appropriate and timely action if necessary.

12. Support and prompt action

After drilling to detail to ascertain the root cause of a notable change or event, users must be enabled with a range of options to share the new information and their associated thoughts with others, in order to drive appropriate resultant action.

Such information collaboration and decision-making options should include, but are certainly not limited to, the ability to:

  • Email the relevant report to pertinent and affected stakeholders
  • Add contextual knowledge to the reports in question via annotations and comments (discussion threads) and have relevant users with access to those reports notified
  • Add decision-widgets to discussion threads to facilitate voting and polling to enable fast and effective collective decision-making
  • Embed the applicable and fully interactive dashboard or report externally to the BI tool, on any third-party Web-based platform, to allow external stakeholders to understand and act on the emergent issue

Where to next?

Remember to take a well-earned break from designing your next dashboard to revisit us for the second part of this BI dashboard best practices blog series.

Top Business Intelligence dashboard design best practices (Part Two) >

Register for our BI dashboard best practices Webinar series >

 

Share This Article
Exit mobile version