Visualising the new world of business data

There are lies, damned lies and statistics. It’s a maxim worth remembering when setting out in the corporate world. Businesses like to consider the secret to an innovation or new idea is hidden deep within the reams of data, charts and presentations they can plot.

However, it’s always important to be wary: truncating an axis line, or exaggerating a trend by zooming into to a small field of data to show dramatic trends where none really exist, are tricks that can easily be played, by accident or design.

With this glut of data making it hard for start-ups and executives to identify a clear path to success, having numeracy around data visualisation is hugely important.

"As the world gets more data-driven and people throw out more and more charts to prove their point, what we're finding is that people are pushing the boundaries of charts more and more to sway public opinion in the world of business and politics. The only way to respond to that is to become more literate in this area," says Scott Berinato, self-professed dataviz geek, Harvard Business Review (HBR) senior editor and author of Good Charts: the HBR Guide to Making Smarter, More Persuasive Data Visualisations.

It isn’t so much a question of the information presented being right or wrong. Context and field of view are often the issues in question and the narratives they can deliver can be misleading.

Plot worldwide sales of vinyl records on a bar chart, for example, and you get a steep curb from 1993 to the present day. Moving the depth of field back to the heyday of vinyl in 1970 and you get a very different story. Vinyl sales remain tiny in historical terms.

Ethics and motivation are key issues here. Charts don’t have to show every data point and the best ones don’t. “There’s an important distinction between accuracy and precision,” says Berinato. “The real point is about obscurantism. Ask yourself: Is someone hiding something?”

The London Underground map is a classic case in point. It is far from being geographically precise but it still functions as a really useful guide as to how to navigate the city. Good slope graphs work in similar ways.

Simplicity

The key to a good chart is simplicity, Berinato observes. Every chart should tell just one story and each piece of information presented needs to work, with all superfluous information excluded.

“Less is more is the principle. You need to communicate one idea clearly rather than presenting numerous ideas pulled from a spreadsheet. Businesspeople like to present lots of information because they fear people will think they have not worked hard enough if they haven’t. Sometimes there’s also a lack of confidence to edit the information down to what is really important, so the trick is to step back and think about the message you want to communicate,” he says.

Berinato says the first key tools that need to be employed are a pen and paper. Ideas need to be jotted and a narrative formed around what is the key message that needs to be presented.

Data visualisation has assumed huge importance for organisations. We live in an era that has seen an explosion in data capture. Making sense of that data is vital.

Berinato says that two well-established principles hold true. The first is expressiveness: the chart says everything you want it to say and doesn’t mislead. The second is effectiveness: the chart is presented in a form that conveys the information most effectively and efficiently.

“They remain the constitutional rules in this area but the devil is in the detail,” he says.

Software tools

There is now a plethora of readily available and cheap (sometimes free) software tools to make chart building much easier. While there is an obvious danger here in giving licence to create boring slides – think bad PowerPoint.

Berinato takes a more optimistic view: “The software can take the drudgery out of the mechanics of chart-making. This can free up time to think really creatively about what you want to show.”

That’s not to say that the design has to be overly elaborate. Executed well, simple bar charts and scatter diagrams work well.

“The design aphorism is: ‘When everything is bold, nothing is,’” says Berinato.

“The other common mistake is including too much text. You need a basic amount of text to support the visual, not compete with it. But too often text just repeats what the graph clearly shows. Our eyes will be drawn to the visual first.”

Technology is having an impact on the world of charts too. The next big wave of improvement in data visualisation will improve interactivity.

“Boardroom presentations will become more of a two-way process. A presenter might be asked by his CEO about how the data shown works fora particular subset, for example. ‘You’ve shown our sales but can we see how our sales are growing for females or for under 25,’ for example. This will be increasingly possible to show in real time.”

Berinato's book is peppered with examples of the good, the bad and the downright ugly in this area with lots of practical tips on how to improve presentations. Charts should no longer be the preserve of data scientists and professional designers, he says – an interesting observation from a specialist who does just that for HBR.

“We can all become more numerate in this area and we are better at this than we think,” he says. “The more we practice it, the better. As the world gets more data-driven and more people use data to make their point, this is becoming more and more important.”

Good Charts, the HBR guide to Making Smarter, More Persuasive Data Visualisations, by Scott Berinato, is published by Harvard Business Review Press.