How to build a dashboard: the 8 key points you shouldn't miss

Data Intelligence

How to build a dashboard: the 8 key points you shouldn't miss

Sonia Moya | Jun 24, 2019

When developing a dashboard on a tool, you’ll normally find yourself with the prior definition used by those who already consume that information, regardless of which tool is used. In many of these cases, it will be a definition inherited from predecessors in your job or, worse still, from the capabilities of the old tool.

However, this shouldn’t stop you from building an effective dashboard or control panel; that’s why I'm outlining a series of questions, the answers to which may lead you to the development of the management tool that will allow you to analyse your organization's key information.

1. Who will be the recipient of the dashboard.

The first thing is to know who is going to analyse the information that you are going to represent in the report you are compiling, since an executive who wants to attain the situation and the development of the company will not have the same information needs as a marketing director who wants to analyse the fluctuation in sales by product, by region etc. The report should be compiled according to the needs of your audience.

2. What information is relevant.

In line with the previous point, depending on who the recipient of the dashboard will be, it will be important to figure out what data you want to analyse. This information will be aligned with the company's strategy and objectives, which will be measurable through the indicators shown on the control panel.

These key indicators should allow the identification of entrepreneurial actions that can drive the organization to fulfil its strategy. When you decide what data you want to show on the dashboard, you should be selective with the amount of information to represent, and leave the possibility of obtaining more detailed listings for more detailed reports. The information on a control panel should be summarized.

3. What terminology is used.

When it comes to analysing information, speaking the same language throughout the company is essential. In this way, the data displayed is understandable and comparable in different departments and areas of the organization, optimizing analysis times and avoiding arguments about what the represented data means.

To achieve this, the information must be correctly labelled using common names to define the varying magnitudes displayed, as well as to name the different axes of analysis of these indicators. It is recommended to explain the metrics and units of measurement, and that these be consistent on all displays. If necessary, the numerical values can also be shown in graphs, preferably reduced to more aggregate units (e.g. units of one thousand or one million, depending on the size of the data).

4. What place each element used should occupy.

When designing and constructing the control panel, the represented elements should be organized in a uniform way that facilitates a quick and simple reading of the data; visualizing the most relevant information and presenting them in an orderly manner, aiming for the utmost clarity of its content. Within the control panel, in addition to data representations, a title must be inserted identifying the content, corporate logo and depending on the timing of the data, the date or month of execution. Then from left to right in order of priority, the graphs with the key indicators to be analysed will be inserted.

5. What type of graphs to use depending on the type of data to be represented.

There are a variety of graphs that can be used to represent information; however, choosing the right graph is critical to getting the right answer. For example, in order to compare data by different characteristics, the best thing to do visually is to use a bar graph; but if you want to compare data by geographical distribution, the use of maps is highly recommended.

If we want to visualize trends and analyse development in time of different metrics, line graphs are an excellent option that can be used in a variety of ways (with double axes, combined with bar graphs, shaded areas etc.) that make the information more visible. Other graphs that can be used are circular or pie charts, heat maps, bubble graphs, scatter plots etc., but always keep in mind which type of graph represents the information to be analysed in the best way and make it more visible.

6. What corporate images and/or colours should be used.

When displaying data within a control panel, the use of colours should always aim to facilitate the visualization of information; therefore, if you choose to represent negative values or falls in profitability in red and, in contrast, positive values or increases in profits, sales etc. in green, this is something that everyone identifies straight away.

At other times, you can choose to represent the values, according to different dimensions of analysis (regions, lines of sales etc.), using a range of corporate colours that are normally used in the company's branding. The use of corporate images, such as the logo, within the dashboards, also contributes to the corporate identity within the organization, so it is advisable to include these elements. The use of these colours and corporate images should not divert attention from the data displayed, which is why bright colours or backgrounds that are too bright are to be avoided.

7. Which elements are useful to optimize the navigation of displayed information.

Once the control panel is implemented and a summary of the main key indicators is obtained, it is normal to want to obtain a greater insight into some of the displayed data; which is why affording the dashboard some dynamism can be a great advantage. This is achieved by incorporating elements such as filters, new breakdown features, skipping to other more detailed reports and even to the system documents where the data we are analysing was taken.

Filters allow us to focus attention on a specific data region, for example, on a specific sales line, on a specific period etc. The breakdown of new characteristics allows us to incorporate more detail into the analysis that is being shown, for example, in the face of a large deviation, to observe whether it has been a specific region, or a specific product that has had the impact on the indicator observed. Skipping to other reports provides a detailed analysis of the elements that have led to a given outcome, but may also explain the origin of a forecast, as defined.

8. What support will be used to visualize the Dashboard.

The way in which the dashboard information will be displayed: whether a mobile device will be used, whether it will be analysed directly from a PC, whether it will be displayed on a big screen, or if on the other hand it is will be printed or attached to some presentation to be sent to shareholders, for example; it also has an influence on how it will be designed. Although there are technological tools capable of mitigating the differences when consuming dashboards on different devices, it is commonly a factor that is taken into account when deciding, for example, the font size or what elements will be used (navigation to other reports, types of graphics etc.).


The answers to all these questions will undoubtedly help you build an ideal dashboard, which will allow the exploitation of information in a way that is useful, homogeneous and consistent for the company. Therefore, the information on the control panel will be reflected in a clear, specific and concise way, an accurate summary of the activities that are monitored over a fixed period - to be as useful as possible for those to whom it is addressed.


Are you ready to make meaning from your data?

From architecture, to analytics and strategy - we can help your organization in any data-related project: view our services and solutions for data intelligence.