In an earlier blog post, we explored the concept of storytelling with data visuals and how important storytelling is in creating a link between the data and the reader. This link creates a personal connection with the data enabling the individual not only to relate to the data but remember it. One of the functions of a data analyst is to communicate what the data is saying. It has been observed that many individuals with advanced degrees in subjects including economics, mathematics, and statistics struggle when it comes to communicating in an effective and efficient manner what the numbers are telling them.
Feldhauser states our slides and reports are getting lost in a world full of bold, impactful, digestible and actionable messages. To get our insights across he suggests we need to think like a marketer.
“We are ninjas at finding insights and rock research design and while our research projects produce marvellous insights …. they don’t make it across the line into the hearts and minds of decision makers.”
Horst Feldhauser
Form and functionality
There is no one way to tell a story. Think of your favorite fairy tale and the various different ways it has been told to you and the different ways you have told it; from the traditional oral sharing of it, the number of different ways it has been written over the years, right through to the Disney version and remakes; it is the same with telling a story with data. As a data analyst, you can shape and present your data in any form so that it makes a connection with the reader and encourages action. Freelance graphic designer Severino Ribecca has developed a helpful catalogue of different data visuals.
In order to communicate ideas and insights to the reader in an effective manner, the visuals have to be both visually appealing but also functional. Vitaly Friedman says that often says often visual designers get the balance between design and function wrong.
“There is no point in creating appealing visuals if they fail to communicate the insights.”
Vitaly Fieldman
According to Feldhaeuser there are generally five levels of visualisations:
- Static infographics
- Infographics with links
- Linked pages with multimedia (videos, audio, graphs)
- Interactive dashboards
- Integrated analysis and visualisation tools
The first three demonstrate and communicate insights while the last two allow the user to explore the data and find insights themselves.
Fernanda Viegas and Martin Wattenberg of CNN propose some “rules of engagement” that can lead to successful visualisations:
- The best kind of visualisation is one you can relate to. Ask yourself, can your readers see themselves in your data and visuals.
- Because visuals are visual, the design is critical. According to Erik Spiekermann by the time the first data point hits the screen you are communicating. Your use of colours, font and other design elements form a strong impression of your intended message.
- Visualisations are social. If it can be shared and discussed it creates more interest. Discussions can also lead to a deeper understanding of the data as people ask questions and discuss interpretations
“…fresh data and infographics spread across Twitter, Facebook and other social channels precisely because they are able to tell a story in a concise, compelling and visually appealing way.”
Dan Trefethen
Design of visuals
There are a number of benefits to using visuals to tell your story. Visuals help you deliver your message by showing rather than telling them and people process images quicker than words. In their Graphic Design Essentials course, HubSpot Academy notes the vast range of tools available to create and add great images including various:
- Stock image sites like iStock and Shutterstock
- Design tools like Canva and PIXLR
A lot of these are available cheaply or free and you don’t need a degree in graphic design to come up with great images.
“Using the right image directly influences your chance of success with your audience.”
The design of your visuals creates and sets the mood according to Viegas and Battenberg. Natural visuals use subdued tones and discreet typesetting a serious mood. Traditional analysts believe that the data should speak for itself and remain neutral sticking to basic tables and graphs. On the other hand, Viegas and Battenberg note that all data sets are not the same and that emotions and a strong voice aren’t sins in other forms of media so shouldn’t be in data visualisations. A good example is that journalists don’t cover causalities of an accident the same way they the sports scores so why should data analysts?
“We can stop hiding behind the idea of neutrality and embrace the power of expressiveness”
Trefethen states that emotional content makes a connection and if you can get your reader to feel something the chances of them connecting with your visual, the data and the story you are telling increases.
Design Elements
There are a number of elements to a visual, these include:
- Colour
- Images
- Typography
- Composition
Colour is very important as it can be used for more than aesthetics. Colour can be used to attract attention as well as convey and provide visual cues to meaning. A classic example is the use of the colour to indicate a positive or red for negative figures.
When using colours in design Hubspot recommend thinking of complementary colours. A trick is to use the colour wheel. Pick the colour you want to use and its complementary colour is directly opposite it on the wheel.
The images and text you use on your visuals are vital. HubSpot advises the image you use, your message and how you word it and even the font you use can mean the difference between reaching your audience or not. People expect to see fonts like Arial or Helvetica that they are familiar with. To improve the readability the use of serif fonts is recommended.
In terms of graphics, it is a good idea to have a look at your organisations’ internal style guides and see if there are any images you can repurpose. This not only saves time but also money. As mentioned earlier there are a number of free or low-cost stock image sites as well as graphic design platforms like Canva and Pixlar that provide basic templates that assist you to create effective graphics.
It is a good practice to test your graphics on various mobile platforms to ensure viewability. According to Statista just over half the world’s website traffic in 2018 was via mobile phones. Images should not appear stretched or pixelated.
In terms of composition you want to keep the design of your visuals simple; two ways of doing this is to use whitespace and remove anything that isn’t necessary. White space declutters the visual assists the reader focus their attention on what you want them to see. Every item in a graphic should serve a pourpose and to keep it simple should only have one pourpose.
“Focus on getting the key things right – colour, imagery topography and composition – and your designs will hit the mark with your audience”
HubSpot
Wrapping up I think the observation from Feldhaueuser is right, it is not the quality of our work or the value of our insights but our delivery which is the problem. His proposed solution to this when it comes to presenting our data and creating visuals we think like a marketer and swap our researcher hat for the marketers hat.
Some final generic tips from Feldhaueuser for successful data visuals:
- Challenge the norm but only if it adds values
- Understand your audience, what drives their decisions?
- Make it clear, get to the point
- Have an engagement factor, something to hook people in to find out more
- Make it memorable, credible and relatable
- Obtain ideas from colleagues and from outside the business. Never stop learning
- Get audience feedback, did they understand your message?
