By Sandra Galiano - "If you don't show it, it doesn't exist" - This is one of the mottos we follow in Business Analytics. In theory, it could be applied to many things. You can be the best at your job or what you do, but if you don't demonstrate it, you don't allow others to see it, and it will never be known.
When we work on a project involving technical concepts like regressions, analysis, data processing, and big data techniques, we gain a deep understanding of how they work and use them to achieve our desired results. For instance, we can predict a company's future, identify reasons for poor product sales, or determine the best time to adjust prices for specific products. These are only a few examples of the countless outcomes that advanced data techniques can offer to companies.
Here we face a problem: the engineer will understand us, your fellow data scientist will understand you and other analysts will understand you, but with the director, executive, and decision-makers, who may lack technical knowledge, we face a challenge.
We can't simply present tables filled with numbers, code scripts, and predictive models during meetings and expect everyone to comprehend.
This is where data visualization comes into play. By creating visual and graphical representations of our insights and conclusions, we can make them understandable to those without data expertise. In doing so, we can effectively communicate the value of our data-driven decisions and ultimately achieve our project goals.
Insights are usually presented in one or several dashboards. These are canvases where information is displayed in a style that mixes infographics and graphical representations.
The dashboards can be adapted to different platforms where the graphics will be displayed, such as mobile screens, desktop screens, etc.
The main advantage of dashboards is their digital nature. Before, if you wanted to present your graphs, you had to print them and explain them. You've probably seen thousands of photos on the internet of a table full of papers with graphs. What problems does this approach entail?
1. The obvious paper and resource consumption
2. It's not efficient: To begin with, in a physical format, less information can be condensed and related than in a digital format, and the most important thing is that it's static.
When we have a dynamic dashboard, we can establish filters and "play" with the screen to see the information we want or more specific information.
Before, if you have a sheet with a graph of, for example, all the products sold in a store in a month by quantity to your superior, and he/she said, "No, but now I want to see the specific sales of this product in this location," you had to go to the computer, edit the graph, print it again, and explain it again. This way, we save not only time and resources but also the person seeing it feels more functional when interacting with the dashboard and seeing how they can find the information.
- Colors: Colors are very important, they must be visually appealing and aesthetic. Most software programs come with pre-made color palettes that combine well and provide a clear view of the information.
- Categories: Categories are preferably each represented by a different color. This way, the legend or label only needs to be placed once. That is, if in Chart 1, blue represents Product X and red represents Product Y, it is ideal only to specify once that red is X and blue is Y, and therefore assume that in the rest of the charts that have the variables X and Y, these colors will be used. There's no need to explain what blue and red represent in each chart.
- Gradations: They are important; sometimes there are no categories but gradations, such as the number of inhabitants on a map. We will represent the higher numbers with darker colors and the lower ones with lighter colors, in a scale. But it is best to use the same color, only changing the darkness or lightness of the color.
- 3D figures should not be used on a dashboard unless essential. The reason is simple: we want to simplify complexity, why make it complicated? If we use 3D graphics, our audience's brain will have to work harder, first converting the image into 2D in their mind, then understanding the representation, and finally interpreting the graph. This is partly why some people get headaches when watching 3D movies because their brain is working on converting the images into 2D, which is what we are used to.
- Filters: We have mentioned them before to make dashboards dynamic. They should be placed on the edges or corners of the dashboard, never in the middle.
- Tree Maps: They are the great hated ones. Care must be taken when using them. They will be helpful when we have few variables that are very different from each other because the distinction would be clear. Still, in terms of percentages, they can make it more difficult to differentiate which square is larger or smaller than another, especially when you move away from the dashboard, as would happen with the audience of a presentation.
- Pie Charts: Categories in pie charts are represented from largest to smallest. The more significant categories are painted from the center to the right, and the smaller ones are left at the end of the circle. This is because the more significant categories are usually the ones people are more interested in seeing at a glance.
💡 Extra tip: It is best to create a public dashboard with important conclusions first, and then, if you want to go deeper, you could give each part you want to focus on its panel.
To choose the appropriate chart, it is essential to consider the type of data being visualized and the objective of the visualization.
● Bar charts: are used to compare numerical values. They help show the distribution of a categorical variable in a data set.
● Line charts: are used to display trends and patterns over time. They help visualize time series data and compare multiple variables.
● Pie charts: are used to show the proportion of each category in a data set, especially when you want to highlight the proportion of each category, but are not recommended if you want to show multiple categories or compare values.
● Scatterplots: are used to show the relationship between two numerical variables. They help show patterns or trends.
● Area charts: are used to show the evolution of a numerical variable over time. They help compare multiple variables on the same chart.
● Histograms: are used to show the distribution of a continuous numerical variable. They help visualize the frequency of each value or range of values in a data set.
● Box plots: are used, for example, to represent actions or results derived from data mining. They are graphs that collect different types of information at a single glance.
Many charts can be used to visualize different types of data. The choice of the appropriate chart will depend on the nature of the data, the objective of the visualization, and the audience to whom the information is being presented.
Multiple data visualization software is available with powerful tools to manage, clean, transform, and analyze data.
The most recognizable and appreciated tools in the field of Business Analytics are:
● Microsoft Power BI: a tool that is also a Business Intelligence solution that provides a wide range of data visualization functions, including pivot tables, charts, and diagrams.
● Tableau: offers a wide range of functions and tools for effectively visualizing data in an easy-to-understand way.
● Google Data Studio: an online data visualization platform that allows you to create customized and interactive reports and dashboards. It also integrates with other Google tools, such as Google Analytics and Google Sheets.
● QlikView: a Business Intelligence tool that offers a wide range of data visualization tools, including pivot tables, charts, and diagrams.
● D3.js: a JavaScript data visualization library that allows you to create customized and highly interactive visualizations.
It is important to note that the field of data visualization is constantly evolving, and staying up-to-date with the latest tools and techniques can give you a competitive edge in the job market. As companies increasingly rely on data-driven insights to inform their decision-making processes, the demand for skilled data professionals who can effectively communicate complex data through visualization will only grow.
Moreover, becoming proficient in data visualization can also benefit your personal career growth. By presenting your data in a visually compelling and informative way, you can effectively communicate your insights to stakeholders and decision-makers, making you a valuable asset to any organization.
So, whether you're a data analyst, scientist, or simply interested in data visualization, investing time and effort in learning these tools and techniques can bring great rewards both professionally and personally.
About the author: Sandra is a tech-savvy young entrepreneur currently President of Generación Empresarial, a youth-oriented university association committed to creating the next generation of successful European business leaders.
By continuing to use the site, you agree to the use of cookies. more information
The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.
