We might expect our data visualisation charts to transform data into insights. But our data is only as good as our audience’s ability to consume and share it. Done right, data visualisation will be effective in enabling analysis and constructing insights. Unfortunately, as anyone who has sat through a chart-heavy PowerPoint presentation knows, many attempts to visualise data can undermine the key message and disorient the end-user instead.
Effective data visualisation tells a clear story guided by a logical data flow. And choosing the right chart type will help you source and convey the story in your data. It is in human nature to focus on the patterns around us - patterns that affect the weather or human behaviour or the design of architecture. We notice patterns because they help us identify trends or outliers that can influence our future wellbeing. Selecting the right chart type will reveal patterns and trends that help our users instantly comprehend the significance of the data set being visualised.
When deciding between data visualisation methods, choose a chart type and formatting that is streamlined and easy for users to understand and analyse. For instance, a line chart makes visualising trends a lot easier than bar charts!
Line charts demonstrate trends quickly and succinctly, in a way that’s hard to misinterpret. In particular, they’re good for comparing changes over time. For businesses, line charts can help easily identify positive and negative revenue trends that can be taken into consideration for future performance. Businesses are able to compare different metrics such as sales by month, over the same period of time.
At first glance, pie charts may seem like the friendly data visualisation choice. They simplify comparison and let the user picture how a few categories compare with one another. But when we move out of math class and into a business setting - the sheer volume of data analysis required makes pie charts much harder to comprehend. Ever tried comparing 20 different pie chart categories all at once?
Simply put, the pie chart is not an ideal option for comparing more than 5 categories. If there are more than 5 categories or if the magnitude of the categories are very close to one other, a bar chart would be a better data visualisation tool in this instance.
Format your charts with the right colours and labels to make them easier to understand and more visually appealing. Used effectively, colours enhance comprehension and draw the eye to important data. Adversely, colours used poorly can confuse end-users and cloud your insights.
An easy way to indicate different values is to vary the intensity of colours used. In the example below, the more intense shade is associated with the category of higher value.
Colour can be used to highlight various categories or represent secondary metrics in your presentation. However, designers should avoid traffic light colours (red, yellow, and green) for normal categories. These colours are mostly used for alerts, negative/positive numbers and thresholds. Bear in mind as well, that users with colour vision deficiency may see the use of red and green together, as brown.
The world of big data is moving towards presenting information in a more accessible manner. Avoid using too many colours unnecessarily. And when you do use colours, apply them with meaning so that they add, rather than detract, from your insights. Consider using colour gradients if there is a need to differentiate more than 5 categories. Gradients are also another way of sorting data.
A common mistake spotted in data visualisation is the underuse of visual elements like colour, graphics and labels. Learn to support data visualisation with labels that display precise figures. Neat and comprehensive data labelling enables users to glean actual values and percentages at a glance. It is especially useful for situations when users lack the time to interact with the chart and view additional data in the tooltips.
Effective data visualisation can provide your analysts and business users with convenient access to data, as well as the means to self-service data discovery and exploration, to bring about improved efficiency. When presenting data, sort the information in ascending or descending order wherever possible, unless you need to sort by time. Sorting can help convey findings effectively, pinpointing, for instance, the best performing category or the worst.
Adding clarity to your charts is also about not misleading users with data presentation. When visualising a dual-axis chart, pay attention to whether the axes can be synchronised to tell your data story more accurately and avoid misinterpretation. As in the example below, you might need to present two different metrics, such as Sales and Profit, simultaneously. Effective data visualisation in this instance, would have you synchronising the scale of the axes to help users interpret the data as it was intended.
In conclusion, applying a user-centered design approach improves the presentation of data.
“We don’t design for the data, we design for humans. Therefore, we have to consider the reader and their context when designing visualisations,” explains data visualisation expert Benjamin Wiedekehr. Never lose sight of your story and direct your audience’s attention to its salient points, simply by applying the best practices above. The next time you approach data visualisation, you’ll find that the proper use of charts, sorting, colours and labels will clarify and simplify your story, and drive your audience to action.