Building the Perfect Dashboard

Friday, March 19, 2021

After few years of presenting a large amount of data to many different types of audience, and after documented me well enough and received so much feedback about my work, allow me to share with you what I consider the most essential key points to design any dashboard and put it in production, so your audience will be able to consume data analytics.

Let me start with the most crucial and fundamental pillar of everything regarding dashboarding: simplicity.

1.    Design with a purpose

Yet, the most difficult point. Start asking yourself ‘why are we doing this dashboard?’ It is the most difficult point to solve, as one dashboard will be consumed by many different individuals and each and every one of them have different needs (in terms of what data to visualize and how to do it).

So, bad news for you as a BI developer: satisfying everybody’s needs is simply impossible.

The challenge here is to globally understand what most of your audience will need and try to satisfy most of their needs, and inform your audience that it is simply impossible to satisfy everybody. So, you will have to do a great exercise understanding what all your customers need and identify the key relevant measures and data.

We all know that as soon as we design and productional a dashboard we are opening ‘Pandora’s box', as everybody will start asking us for a new view, a new perspective, new colors, new chart, more filters, etc etc etc… As a developer, you must prioritize what to show and how to show to maximize the utility of the dashboard

2.    Important things always at first sight

A dashboard is not a report. Keep that in mind, and make sure your customers understand this.

Design your dashboard focusing on what is important. Empathize with your customers. Keep in mind that the majority of your customers have been analyzing data using spreadsheets and they are very used to bring many data points at once in the same spreadsheet. You shouldn’t do that

The purpose of a dashboard is to summarize the most relevant data points into a single view. Not everybody will have the same opinion about ‘what is relevant or important to view first, but then you may now understand what has been described in the previous point.

3.    Simplicity, Simplicity, Simplicity

This is maybe the central mantra. Power BI, as well as most of the visualization tools at the market, offer a big number of functionalities, visualization, formatting, and many other features. Which is fine. However, keep things very simple.

In addition, consider who is your audience. If your audience is not mature with modern visualization tools -perhaps because you just recently automated manual reporting-, do not overwhelm the audience with fancy visuals and strong color. Try to keep them very simple but functional. As soon as your audience becomes more familiar with the dashboard and its visualization, it would be a great moment for you as a developer to also improve visualization by adding new functionality or more sophisticated visualization.

Your audience will appreciate that a dashboard is a ‘living entity’ that evolves with their needs.

4.    Be consistent

When we are developing dashboards, we as developers create a lot of mess. We change visuals, colors, formats, alignment, etcetera. Make sure that you have a standard for your work once you finish.

Bear in mind that different people respond differently to the same stimulus. In a dashboard, we work with colors, shapes, figures, alignment, etcetera, in addition to data and analysis.

Be careful selecting the right color palette, the standard shapes, text size, fonts, and other standards acceptable by your customers/organization that would be consuming your dashboard. In fact, this is maybe one of the most important aspects of a dashboard when is in front of your customers.

5.    Show the context

Finally, perhaps one of the most important aspects -and perhaps one of the most ignored aspects- in dashboarding is: provide the context to the data.

Looking at a KPI isolated is fine for some individuals but understanding if the value of such KPI is relevant or not is a different story.

Make sure that you add the global context surrounded that KPI, so not just one or two customers will find it relevant, also the rest of your audience.  

If there is a KPI that shows ‘$10M in Revenue’… ask yourself, why is this important? Is this too much, too little, as expected? Well, consider adding a line chart showing the historical trend, so you are adding context to that KPI.

On the other hand, avoid aggregating ‘noise’ to your dashboard. By noise, we mean data that is not related or relevant to the data that you are meant to show to your customers.

Find below a great example of what is NOT a good dashboard

Find below an example of a good dashboard


More DAC Content by Pablo Moreno

How to Interview a Data Scientist

Roadmap to Digital Transformation

Excel or Power BI? ... It's Not About the Tool

Text Analysis with Power BI

Follow Pablo Moreno's DAC Profile HERE

Discussion

There are no comment yet, be the first to ask!