Your company has spent a lot of resources on their new Business Intelligence tool, and now it is time to start exploring the data and looking for all those valuable insights that are supposed to justify the high price tag. When approaching visualization, it is important to move beyond showing trends and KPI’s. At Cervello, we follow an agile approach for our clients that includes Value Based Design. Value Based Design emphasizes delivering real business value by not only showing the key value indicators (or KVI’s), but also creating a series of dashboards illustrating the drivers, action points, and decisions that build up to each KVI.

Here are five key principles that we think you should keep in mind, no matter what visualization tool you are using:

 

1. Tell your story at-a-glance

An effective dashboard uses layouts and visualizations that tell a clear story that can be easily interpreted at-a-glance. To get to that at-a-glance state, it’s important first to make sure that you know your audience and that they know the data. For instance, you will want to make sure that you are using organizationally common terms for measures and attributes. Also, it’s best to try to avoid acronyms that might be unclear, but also don’t be so verbose that the dashboard becomes too text-dense.

A great test for testing the “glance factor” is the squint test. A user who is familiar with the data should be able to get the gist of the dashboard even while squinting at it which causes smaller bits of information to blur out leaving only the “big picture”. Testing your dashboard for the “big picture” is a way to ensure that your visuals are telling your story without added explanation. If your choice of visuals, text, and layout don’t pass the squint test, you will want to reconsider the approach you’ve taken, and rework it so that your story on a particular dashboard is always clear and never requires explanation to an informed user.

Your dashboards should tell a story which flows naturally from dashboard to dashboard so that a sense of the whole story is easy to discern from the disparate parts. Help users put themselves into your stories by making them very clear, erring on the side of simplicity. Remember, dashboards should not be data dumps that force the user to construct their own stories. Great dashboard designs are like a film. They need to be edited together in a narrative that is clear and has continuity between scenes. The goal of your “film”, however, is not to entertain. Its purpose is to reveal patterns and anomalies to help management make informed and confident decisions, ask intelligent questions and take corrective action when necessary.

 

 

analytics-effortless

2. Make data effortless to consume

Simplicity and clarity are key. Sometimes less is more impactful. Ask yourself, “Is this meaningful”? Don’t focus on flashy bells and whistles. Instead, concentrate on the pure story your numbers should be telling. The intent of a well-designed dashboard should be for it to only take a moment for the user to comprehend the context. You want an end user’s experience with your dashboard to be as frictionless as possible.

To make the user’s experience frictionless you need to understand how they will be consuming the data. Get data in front of the users as quickly as possible. As you continue to iterate and refine your dashboards, you will be able to better understand the key areas of focus. Presenting those areas first with as few clicks as possible allows a user to consume data more effortlessly. Try to remove as many steps as possible between the user opening the dashboard and viewing meaningful results. Minimizing the need to select business units, geographic regions and defining time periods helps speed up user consumption of data with the fewest number of clicks. All of these different controls can be very useful to explore the data, but users should not have to navigate through multiple screens of options before ever seeing meaningful data (not just data).

A great way to ensure a focus on the essentials is the use of Flat Design, a visual design methodology pioneered for mobile devices. It eschews extraneous visual effects like bevels, shadows, 3D renderings, isometric angles, etc., and focuses instead on typography and content, with large, untextured expanses of color to create groupings on as simple of a level as possible. Flat design ensures that the intent of a well-designed dashboard only takes a moment to comprehend. Overall, the intent of flat design is to ensure a frictionless consumption experience where the least amount of visual information possible supports the user and the story of your analytics content.

 

 

analytics-navigation

3. Make your dashboard collection easy to navigate

You should always strive to avoid novel and overly-imaginative navigation: don’t underestimate tried and true user-tested navigation styles! You shouldn’t reinvent the wheel when it comes to user interface expectations and effort because browsing, clicking, dragging, and drilling are very ingrained user habits. A good dashboard requires little “figuring out” and should be as self-explanatory as possible. Avoid cramming too many visualizations onto a single dashboard. Instead, let the story be told using as many dashboards as necessary, like scenes from a movie so that each dashboard and dashlet is self-explanatory.

One of the most important design considerations is to use a responsive layout. You may have to forgo some of the precise pixel-perfect formatting you’ve been used to in the past, but it will be worth it. Choosing to design for “mobile first” means a focus on content, context, and continuity across devices and platforms. Responsive layouts naturally constrain design choices around content and eliminate inefficient use of resources focused on subject aesthetics. Responsive layouts are inherently a “content first” technology which has many upsides and few downsides. Responsive design helps to keep things clean on all different sizes of screens. It is a best practice to try to remove as many images, menus, buttons and text as possible, and use zooming and drill-down capabilities instead. Ensure that the few navigation icons you do use are meaningful and label if necessary. Selectors should be clear and concise with only relevant options presented. Since you already have some idea of how your users will consume the data, don’t present options that will probably never be used and just add clutter.

With today’s data visualization tools, there are so many ways to present data it can be overwhelming. Don’t get so excited with all the different options that you lose sight of the goal of each dashboard. Unique visualizations can be valuable when used correctly to expose movement and trends in the data that might not stand out in a more traditional format. Just make sure your users know what they are looking at.

 

 

analytics-roi

4. Understand the ROI for clear design

Actionable intelligence can have an immediate effect on the bottom line. If an insight is obscured, opportunities might be missed, and problems might be overlooked. It’s critical to know the exact user profile and the most important question the user needs answered on a dashboard by dashboard basis. When evaluating ROI, the first place to start is looking at the current way(s) your users consume data (probably a diverse array of systems and excel files) versus how they will consume it on the dashboards.

ROI is where a value based design approach really shines. Instead of just providing high-level information to users you can demonstrate the inputs that make up each of those numbers. By working their way down, users can see where they can actually affect outcomes, the goal always being to guide the user to make concrete decisions. Making a few changes that affect some high-dollar outcomes can quickly make a business intelligence product well worth the investment!

 

 

analytics-color

5. Use a consistent and minimal color scheme

Color should be used very carefully and often carry meaning where no meaning is intended. Focus on a simple color scheme and the use of neutral colors so that when something does need to be called out and highlighted with color, it will be more effective against a generally constrained and muted palette.

Colors of different intensities can be used to direct attention around the dashboard. If all colors are the same intensity, they all stand out as equally important. Similar data sets should use similar colors in the same “family” of different intensities to draw the user around the dashboard. Data sets that contrast should use bold, contrasting colors. When choosing the different base colors to use, look for colors that complement each other and avoid colors that blur or clash. Utilizing your company’s corporate color palette is often times a good idea. A couple of resources to help you get started with palette choice and optimization are http://paletton.com and http://colorschemedesigner.com.

In conclusion
Keeping these five points in mind will help you be well on your way to dashboard designs that are beautiful, clear, meaningful, and effortless to consume while providing measurable value to your end users.

Interested in learning more? Join our mailing list!

Author:

Doug Bonneville