Where Infographics Are Going

At their best, information graphics can be informative, exciting, and attractive. At their worst, they can be misleading, overdesigned, and empty. Infographics are still in their infancy, with a lot of untapped potential. Ideas from visualization can help figure out a future that is much more exciting.

Imagine a slider to control the amount of decoration in an infographic: 0 for no decoration, 1 for the maximum amount possible. Most current infographics are close to 1, with a lot decoration applied. For an infographic to be interesting and compelling, there is clearly a minimum of decoration that is necessary. However, I believe that much of it is gratuitous and unnecessary, and not only distracts from the message, but undermines it.

Let's look at the other end of the scale: no decoration. This is where visualization sits. In fact, I think that data visualization is a true subset of infographics. What else is a visualization than a graphic that informs? And if it doesn't inform, how can it be a visualization? In addition to being bare, though, visualizations are also general, i.e., they do not care what data they depict.

Priorities

As Jakob Jochmann points out, a lot of focus in making infographics is currently on attracting clicks and eyeballs, also known as "SEO juice." That's really too bad, because that shifts the priorities away from being informative and towards shiny, colorful, and maybe controversial, but not useful.

When you look at the work of people like Nigel Holmes, the difference in priorities becomes obvious. You can argue whether he is dressing his graphics up too much, but there is never an attempt to hide the shallowness of the message or the lack of data behind pretty pictures. The message is front and center, reinforced by the metaphors and graphical treatment. In a sense, that is quite efficient: there is ample decoration, but it serves a clear purpose. A lot of infographics you see on the web these days not only have lots of decoration, but useless decoration.

In addition, there is also an underlying dishonesty: infographics often only serve as click-bait for people surfing the web while they're bored. They have no intent to inform, and thus have no value to the viewer. They're ads, plain and simple.

Interaction

Infographics are not only often shallow, but almost always static. You can usually get all the information from them with a few glances, and then you're done. Sure, there may be some clever little graphical jokes hidden, but you're not going to learn more about what the story is actually about.

I envision a future of infographics that go well beyond the handful of numbers from questionable sources that are currently the norm. How about interactive infographics? Why can't the graphical flourishes lure me into some interesting little story, but then provide tons of background data for me to explore? What about showing alternative explanations? Why not let me focus in on a particular area or data range of interest (like that amazing yet simple New York Times visualization of the jobless rate)? What about providing me with the underlying data so I can play with it myself, rather than hiding behind some nondescript "data sources" in tiny fonts on the bottom?

For graphics to be truly informative, they have to be interactive. There is nothing like being able to ask your own questions if you want to really understand something. A particularly interesting approach is Bret Victor's KillMath project, especially his explorable explanations. His focus is on math rather than on data, but I don't see why the same ideas can't be applied to data. Spend some time with his interactive widgets and then tell me that today's infographics are adequate. You will never look at static depictions of data the same way again.

Visualization has developed the tools for rich exploration and interaction. So far, the focus is still largely on exploration and analysis. But there is a huge opportunity to apply these to the much broader area of informing people by presenting data. Visualization alone is not the solution, but neither are shiny, colorful graphics. The sweet spot is somewhere in the middle, and it will take the combined smarts of visualization people and designers to find it.

Posted by Robert Kosara on January 26, 2012.

---

Comments

Jon Peltier says…

"Provide me with the underlying data..."

Yes, please.

Peter Wang says…

Interactive visualization is absolutely the future. Most people, sadly, have never seen their power, and most toolkits today are architected to be one-way flows of a data space -> screen space pipeline. Whenever I showed off Chaco's interactive demos to people, their eyes lit up and they immediately "got it". But most people, even if they can imagine such fluid interactions, have trouble comprehending the immediacy and subconscious understanding that interactive graphics provide, until they use them on a real problem or dataset in their domain.

In order for interactive visualization to become a tool within reach of average designers and data researchers, better toolkits (and, dare I say, languages?) for visual specification have to be designed. Even something as new as VizQL, for instance, is designed to express a one-way transformation of data. The linkages between data is implied in the schema of its data sources, and perhaps some amount of interactivity could be automatically introduced into the plots as a result of this (c.f. Spotfire), but I believe that both Tableau's and Spotfire's approaches are fundamentally limited because they treat the graphics specification language as a separate domain than the data transformation language.

Len De Groot says…

Agreed that interactive infographics are important but they have been in production for at least a decade. A site I contributed to and eventually managed has hundreds. The sad part is that even though the site was started around 2000, the staff that contributed the content peaked at about 10 and is now limited to one person who's primary duties are print-based. That says less about its value and more about corporate journalism's lack of vision. If you want to dance through the graveyard, check out TheEdge.

Personally, I would argue the democratization of data is the future, meaning data is presented in ways that average people can consume. That includes interactive data visualization and things that haven't yet been conceived. My question for the group is, "Is the open data movement meaningless if only us propeller-heads and decision makers pay attention?"

Peter Wang says…

"Is the open data movement meaningless if only us propeller-heads and decision makers pay attention?"

Yes. Information graphics is a subcategory of communication. Visual communications is oftentimes the only way to make sense of large size or high dimensional datasets. If people are to be truly literate with infographics, they need to be able to read and write them - which is why I made the earlier point that new languages are needed for expressing interactive graphics.

Jorge Camoes says…

When I was a young intellectual I loved to discuss how terrible easy listening music was. The music didn't change, but I did. I'm not young or intellectual anymore, and I have better ways to spend my time.

The same with infographics. We spend way too much time discussing them. Marketeers know that a cheap infographic can get traffic and they will use them ad nauseum. Then they'll move on to the next goose with the golden eggs. It's in their nature. Bad music and bad infographics will always have their audience as long as it can be monetized.

What can the data visualization community do? It can try to improve awareness and graphicacy skills but I don't see that. Could replace the hundreds of bad charts in the Wikipedia but they are there, waiting. Yes, the high-end is sexy and I love a well-designed infographic, but it doesn't change much at the bottom.

And yes, interaction is the future. Bertin said that in 1967.

Keith Suckling says…

Good points from all...

The flashy/ shallow Infographic, the simple analytical tool, the DIY data visualiser and the raw data aren't meant for the same people. Why should we be surprised when each of those outputs don't land effectively with a wide audience. Each of them was only ever intended for a specific target group of people.

Isn't the future is personalisation? And so why should this field and its outputs be any different?

Len De Groot says…

Nice point.

Mike Wirth says…

I agree that Interactive Visualization is part of the future, but would add that we need to learn how to make visualizations more usable. We must also not forget the power of animation on visualization. How data moves will soon become a part of the importance of how it looks. Look not just to interaction designers, but filmmakers, too.