5 Tips For Improved Data Visualisation

Data Visualisation

It is often said that a picture is worth a thousand words and no saying remains truer in the world of business intelligence and data visualisation. Used properly, visualizing rows of data via graphs and charts can be an effective method for extracting meaning and therefore critical business insights for companies. This is increasingly important as companies try and find ways to actually harness real meaning from their data and get effective insights on the best ways to improve the operations of their businesses.

The good news is that visualisation technologies have improved to correspond with other data trends. Although charts have been used by companies for decades, simply throwing data into a chart isn’t enough to get true business insights. Instead, organisations are making moves towards using more detailed, interactive charts that mean users can spot outliers, patterns, trends, and even correlations in complex data in a way they never could before.

However, for an organisation to truly experience the benefits of cutting-edge data visualisation, we think there are five things they should focus on if they want to make sure that these platforms are delivering maximum impact for their business.

1. Be Selective Of What You’re Trying To Achieve

Often the first mistake for data visualisation teams is that they cast the net too wide. They want their graphics to represent too much information and, as a result, the visualisation becomes too crowded. A good visualisation represents a small number of measures – preferably no more than seven to nine, as that is all that users can cover at once.

Similarly, be specific with the number of KPIs you want represented in the dashboard – again, keep it to nine or less. After all, too many indicators are distracting. Keep the visualisation simple. The less there is to interpret, the easier it is for the user to understand. If your visual looks cluttered, try a different format. The cleanest format is usually the best.

Another principle is to make sure the format of the information is appropriate for the use case and the user. Chart formats, for example, come in many types and are often preferred by business users. Edward Tufte, a leading academic, talks a lot about “chart junk” which is the complexity that people sometimes put onto a visualisation that can obscure the meaning. So when you create charts, use as little ink or pixels as possible to maximise the data.

2. Choose An Appropriate Chart Type For Data

Be aware of which type of chart you should use and which one will best display the data and also fit your audience who will be looking at it. We’re all familiar with line charts, area charts and pie charts but sometimes certain visualisations are not appropriate.

Pie charts for example are a common format for displaying information, but they are controversial in the world of visualisation. It is extremely difficult to compare one pie chart to another yet they are used in that way all the time. Use them, but use them appropriately, and only to show a relatively small series of data. Become familiar with the wide range of visual tools at your disposal and then choose them wisely based on the overall desired outcome.

3. Using Colour & Perception

The use of colour and perception is very important in visualisation and many organisations misuse it. Using colour well can enhance and clarify a chart so that the users can get the highlights of the data. Colour used poorly will confuse the user and obscure the data. For example, you can use colour to highlight positives or negatives or show how the data changes over time or with different variables.

However, remember when using colours that analysis should always come first. Many companies fall into the trap of using their brand colours in visualisations but these are often not the best option. Businesses should consider what messages and insights the colouring is trying to convey and make an appropriate choice based on this rather than what the branding department has to say. Finally colour blindness is a consideration; don’t rule these people out as users. You shouldn’t use colour alone to transmit meaning if at all possible. Use shapes or use appropriate colours that most everyone can see.

4. Choose Your Data Sets Carefully

Great visualisations start with great data and will only be as useful as the quality of the data it represents. Sometimes when teams receive unexpected results from their visualisations they can’t understand why and their BI tools can end up taking the blame. However, this shouldn’t be the case. Instead, visualisation tools should be used to help spot these issues early so they can be corrected in time to not affect the entire project. This can often be helped from an end-user perspective by being able to identify the difference between an unexpected discovery and a data issue.

5. Just Try It

And finally, just try it. Let your users interact with the visuals where there is a pressing need for it, perhaps in a department with geographic requirements.

Overall, data visualisation has the capacity to really turn around the way your organisation is using – and benefiting from – its data, but only if it’s being used in the most effective way possible. Napoleon is right – a picture is worth a thousand words – but only if that picture is displayed to help you understand a situation and act accordingly.

Patrik Lundblad

As a product manager at Qlik, Patrik Lundblad is responsible for the vision, strategy and roadmap of the product. This involves identifying the needs of the market and to deliver solutions that meets their requirements. Patrik is also involved in the launch of Qlik’s new products to the market and communicating with customers and prospective customers on Qlik’s vision. Patrik has a background as a PhD Student at the Center for Visual Analytics at Linköping University. In May 2013, he acquired his Doctoral Degree with the dissertation Applied Geovisual Analytics and Storytelling. He has excellent presentation skills and is a highly appreciated lecturer in visualisation and speaker at conferences, customer meetings and training sessions.