Author Rudyard Kipling once said that “If history were taught in the form of stories, it would never be forgotten”. Storytelling and data often go hand in hand, but it’s not always easy to get started without a data visualisation guide. With the amount of data we currently produce (some 2.5 quintillion bytes of data every day) making heads or tails of data is almost impossible without analysis and data visualisation.
Many organisations spend time building out their data functions with data engineers, analysts and data scientists. Too often, data visualisation is treated as a bit of an afterthought. It’s relegated to a series of dashboards and self-service data visualisation tools. That’s an injustice. It fails to show the full potential and value of data visualisation.
90% of the world’s data was generated in the last two years alone. That growth will only accelerate as technology such as the Internet of Things (IoT) gains momentum. But, we’re sat on a mountain of data with no real way to extract meaning from it. Worse still, what little meaning we do get from our data is wrapped up in reports and mind-numbing presentations.
Analysis gives you insights on your data, but it’s still just numbers. To really get people to act on your data, wrap it up in data visualisation that drives engagement with the insights.
Data visualisation offers a way to unpick the nuggets in your data. It presents information in an engaging and understandable way across your organisation. If you’re presented with a 10-page report on your marketing effectiveness or an interesting graphic, which one is going to grab your attention?
The success of a data strategy hinges on a data-first culture. You cannot achieve this if your data insights aren’t communicated to your employees in a way that resonates with them individually. Data visualisation is the best way to disseminate information across your organisation. It’s critical to gaining buy-in from your wider team. You can lead from the top, but if individuals across your teams aren’t invested in what you’re trying to do, your data project will inevitably fail.
Many organisations understand the need for data visualisation. The market for data visualisation tools is set to grow 9% year-on-year to reach $6.99 billion by the end of 2022.
But many seem to see a data visualisation tool as the be-all and end-all of visualisation. Investing in dashboards and self-service tools can be useful for regular reporting. However, they are incredibly limited in the visuals that they can create and can become repetitive. Plus, giving the tools over to your employees with little-to-no training in data visualisation fundamentals will fast-track you to ugly, illegible (and in some cases dangerous) visuals.
The foundations of good data visualisation lie in the data. If your graphics don’t represent what your data is saying then they are useless. You need to understand which charts work best with different data sets. Handily there are many chart selectors available to give a steer on this.
Pie charts are a good example of misused visualisation. There’s a heavy reliance on them across many industries. But our brains don’t interpret them very well. We cannot accurately measure the areas of different pie slices or its angles. In other words, the information portrayed cannot be understood at-a-glance, something that is critical to good visualisation.
Many visuals can be functional but not very interesting to look at. To encourage people to use data in their everyday work, and to understand the insights, make the visual something that they want to look at. David McCandless is a master of this. His book and associated website, Information is Beautiful shows many examples of how data visualisation can bring data to life.
However, it can be easy to get carried away with the graphics and forget the information you’re trying to convey. Indeed, a survey of business leaders recently found a disdain for infographics, because they too often prioritised the graphic over information.
The lesson here? Make sure every element of your visual is needed.
That doesn’t mean you can’t be creative. Mountains Out of Molehills, for example, shows a great balance between looks and practicality. It plays on the title with a mountain-like landscape depicting various media scares over the years.
Again this comes down to understanding. Think of your audience’s expertise level. Also, consider the context – something you share with peers in the boardroom might look different to a weekly report or an infographic shared with the media.
Prioritise the information your audience wants to see. Don’t try to communicate too much information or you will lose the underlying message.
Finally, understand what stories to tell. Data visualisation can be a powerful tool to tell the right story, or the wrong one. If you consider the visual Iraq’s Bloody Toll (below), it shows a graphic depiction of the deaths in the region. There’s a dramatic story of mass death that has an emotive impact on the viewer. But if you flip it the other way you will see a different story told. Deaths have been in decline over the past few years. Choose your visuals wisely and always consider the perception that you are creating.
A dedicated data visualiser has the fundamental knowledge to assist with all these elements. There’s a specific skill required to visualise: telling a story with data. A data visualiser needs to be part data scientist, part storyteller and part (UX) designer.
Even if you have limited resources, hiring a visualiser to set you up and ensure you’re following the right visualisation principles will pay off in the long run.
Visualisation provides the context for all your decision-making. You can distort data with the wrong visualisation. Consider Anscombe’s Quartet. It uses the same data with the same statistical properties. But when graphed, each graphic tells a different story. It ties into my last point – making sure perception is accurate.
True data visualisers understand that it’s the story is the star. With data being the leading lady. If you choose the right visual that aligns with your data, then you cannot go wrong. Ensure it’s the right level for your audience and not cluttered. Always start with the data, not the other way around. By letting your data tell its story, you’ll find an effective visual that’s perfectly in sync with it.
Cynozure research calls for stronger data leadership in US financial firms to unlock data benefits
A CLEAR™ Path: How to Foster a Robust Data Culture by Leveraging Concepts from Industrial and Organizational Psychology
Outsmarting the Bots on demand: A practical application of machine learning