When you make changes to a data graphic to give it more ‘wow’, you do so because you are trying to grab your readers attention. That’s fine, to a point. And it can serve the noble aim of trying to get your audience to spend more of their precious attention learning from your graphic whatever it is it has to teach.
However, often those changes to form, the tarting up, also have a degrading quality its function. That is, its ability to clearly communicate its message.
Conceptually, imagine that we could measure the ‘attention-grab-iness’ of your graphic in terms of the amount of time someone is willing to spend studying it. Let’s call this the willing-attention-time. The ‘clarity-of-communication’ we could measure in terms of the amount of time someone needs to spend studying it in order to understand the message. Let’s call this the required-study-time.
A successful graphic clearly needs willing-attention-time to exceed required-study-time.
Now consider that your data graphic takes place in a context. It makes a difference whether it is going out on Twitter aimed at Joe Public or is part of a 200-page report for a specialist technical audience. Think of this context as adding a constraint, a maximum bound, on the willing-attention-time. On Twitter that you’ve probably only got a couple of seconds of someone’s attention, in an academic paper, an info-graphic poster or a data visualisation hard-copy book, you can assume it’s quite a bit longer.
There is also a constraint on required-study-time: the complexity of the message your graphic is trying to convey. If your message is “this company’s share price is rising faster than the market”, that is something you should easily be able to convey graphically in less than a second. But if you are trying to convey how the company’s share price has been driven up through a combination of its position in a growing market, a recent successful marketing campaign, the misfortunes of a main rival and changes in exchange rates… Well, that is going to require more attention from your consumer, more of their time to digest, no matter how well you design the graphic.
Let’s look at some examples. The Economist magazine has a deservedly good reputation for producing good data graphics. Below are two recent examples, relating to the same published article. (Click an image to see the originals in context.) One was posted on Twitter, the other embedded within the article itself. Can you guess which is which? There are no prizes.
The Economist often produces graphics which work well in both mediums, but here the complexity of the message, in dual-chart form, is just too much to put into the context of Twitter. And so for Twitter they have greatly simplified the message and form.
Remember you are the designer. You hold the reins of these four horsemen of data graphics: the form, the function, the message complexity and the context. You will rarely have full control over all of them, but with discretion and influence over at least (any) two of the four, you should be able to create a data graphic with the power to communicate effectively.