Data is at the heart of all discussions about the future of our societies. Back in 2017, Jeff Fochtman, Head of Marketing at Seagate, announced: What’s surprising is not that the production of data to be stored is increasing, but the frantic pace of this increase.
Data production is now at an all-time high and its growth is not likely to slow down (connected objects, voice assistants, autonomous cars, artificial intelligence, smart cities...are all technologies that will increase this data flow). These reasons make data visualisation a key issue for professionals.
Definition: what is data visualisation ?
Data visualization (data viz or graphic representation) consists in visually structuring collected and stored data. This makes it easier to use the data.
Each organization has data. Small, medium-sized, large companies, public institutions and even associations all have data thanks to their archives and newly collected data. These archived data, once digitized, turn out to be real gold mines once their data visualizations are completed.
As you will have understood, data visualisation is the art and the way to transform data into a formidable analysis tool. By showing the invisible, data visualization facilitates and accelerates decision making. It is a valuable tool that is more efficient than simple Excel tables. Thanks to data visualisation, you get to the heart of the matter! Data visualization simplifies the dissemination of information. It provides points of comparison and analysis on trends. It then refines predictions on future trends.
First of all, a good data visualization takes its target audience into consideration. The information presented must be consistent with the duties and time available to your public. It is therefore necessary to get to the point and keep consistency in your ideas. In France, in companies with more than 500 employees, managers would spend nearly 6 weeks in meetings. At the European level 14% of executives fall asleep during endless PowerPoint presentations. Here are some best practices that will help you create better data visualizations!
Now, let’s see how to make a data visualization!
Form is essential in data visualisation
The type of graph is a criterion of choice in data visualization. We do not represent a change over time as we would for a comparison. Here are some examples to use for your data viz.
- Time Change: Line Charts, Bar Charts, Stacked Bar Charts, Candlestick Charts, Pie Charts, Timelines, Horizon Charts, Waterfall Charts.
- Comparison: Bar Charts, Grouped Bar Charts, Bubble Charts, Multi-Line Charts, Parallel Coordinate Charts, Bullet Charts
- Classification: ordered bar diagrams, ordered column diagrams, parallel coordinate diagrams
- Distribution: histograms, box-and-whisker plots, violin charts, density plots
- Correlation: Scatter plots, bubble diagrams, column and line plots, heat maps.
However, the appearance of a graph changes the way we perceive the data. A line graph shows trends and fluctuations in a fraction of a second. A pie chart is useful for quick comparisons of data but is not recommended for accuracy. Moreover, angles and arcs are difficult to assess by the human eye, and the evaluation of the data is disturbed. So choose the chart that best tells your story.
In addition, avoid anything that does not add information to your data visualization. Texts, illustrations and 3D, if they are superfluous, have no place! Also think about prioritizing your information. So, on a bar chart, place the highest values on the left as in the example below.
Colors and data visualization
You certainly know that colors affect our emotions! They affect our decision-making. To convince people, do not neglect the choice of your colors! One of the basic rules in data visualisation is the use of a single color to represent the same type of data.
The use of a large number of colours is not prohibited but is not recommended. Your palette should be limited and your colours should add information to your presentation. Do not use totally different colors. They make your data visualizations unsightly but mainly participate in an unconscious hierarchy of information while blurring the perception of data. Tools such as Colorgorical or Colobrewer will certainly help you to choose your palette.
The colors provide information on quantities. They make it possible to highlight certain data. They also provide information on the value of data sets (hazard levels, positivity, negativity, etc.). The colors have a symbolism that must be taken into account!
Other rules of practice are crucial. Your data visualizations are studied via screens or projectors, bright colors (even fluorescent) will be perceived as aggressive and will fail to understand your presentations. We mentioned above that your targets must be taken into consideration. If your audience includes people with visual impairments (color blindness, photophobia, blindness...) use pastel colors and fillings in the form of diagonal lines (note that this type of filling is to be avoided in case there is no disability).
As for the shapes, your colors must be as clean as possible. They contribute to a better understanding of your data visualizations.
Typography and text for data visualization
The choice of your typography is crucial. Modern and unparalleled typographies such as Roboto, Open sans, Helvetica, Lato and Montserrat are perfectly suited for reading on screens. You will find all these typographies within Powerslide. That said, some good textual practices will facilitate the understanding of your data visualizations. Again, you will have to consider your audience and make your texts as inclusive as possible.
Always add a title to your charts. This title helps contextualize your data visualization and brings precision to it. However, do not have a long or complicated title, be concise.
Write in the present tense. The present tense is a descriptive tense that involves the audience.
Text has a secondary place in data visualisation so it must be kept to a minimum. Your audience should not be dispersed, it should be focused on the highlighted data.
Data visualization makes the world easier to understand. It is a great tool for analysis and projection into the future. It quickly established itself as one of the cornerstones of decision-making. Today, it is becoming accessible to as many people as possible. It is therefore essential to have an efficient data visualisation solution.
How to make a good data visualization? Here is a brief summary:
- Take your audience into consideration (function, time available, disability, etc.)
- Choose the chart that best suits your needs
- Do not insert superfluous elements (3D, flashing, etc.)
- Text and images should be used sparingly
- Choose your colours carefully
- Use modern typography
- Add titles, and captions if necessary
- Write in the present tense
- Get to the point
Edward Tufte (one of the pioneers of data visualisation) draws a link between this search for the essential and the minimalism of contemporary art. Some artists have made it the heart of their activities.