Can a graphic be SO bad it starts to become good?

I feel compelled to share and briefly discuss this, even though it is getting plenty of column inches already via Twitter/blogs. It is an infographic published on and produced for Microsoft to promote their cloud computing in the Asia Pacific.

The accompanying description suggests…

Microsoft wanted to create a visual statistic on how people think about Cloud Computing in Asia Pacific. We created infographic with a flow and simplified the data to make it easy to understand.

I’m having a very strange experience with this particular graphic. Rarely has one piece of visualisation or infographic design ever broken so many established practices, principles and rules of effective communication. Take a look at the balloon element below. Case closed. It is essentially impenetrable on an interpretation level – it clearly does NOT make it easy to understand.

And yet I have probably spent more time staring at it and studying it than most pieces I’ve seen for a long time. Its proving to be a challenge similar in nature to Where’s Wally, I know there is something there that will inform me, but I just can’t seem to find a way of getting there.

So, to answer my rather sensationalist opening question, is it so bad that at some point it starts to become good?

As with any piece of visualisation/graphic work, it depends entirely on the objective. I can’t tell you much about Microsoft’s progress in the Asia Pacific or the attitudes of people in that region towards cloud computing, but the basic concept and connection of Microsoft to cloud computing is now firmly implanted in my mind, much more so than before, because I’ve spent so damn long looking at it. If the objective had been around inherent advertisement the graphic it probably has been good at achieving what it set out.

However, if we refer back to the text that came with the piece, the objective implied from this description is that the graphic represents a visual statistic with data that is simplified to make it easy to understand. In that sense it completely and utterly fails on every level.

Anyway, here’s a challenge, which can you complete first – useful insights from the Microsoft graphic or find Wally in the image below? On your marks, get set, go…


Andy CotgreaveAugust 25th, 2011 at 9:13 am

This post is excellent and represents my thoughts exactly. When I first saw it, I thought, “‘WTF? A typical failure”. However, I then ended up poring over most of the details of the poster. What you’ve done with the Where’s Wally image is capture why the MS graphic is a good one. I felt like an 8 yr old boy, looking at one of my “How things work” books again – sure, the info’s not following “good practice”, but it’s fun to explore, and pretty.

While many other infographics try to achieve “fun to explore”, they fail because they are appealing to the design cliches of the hipster geek crowd. The MS one appeals to the child and succeeds, against all the odds.

I figured that if I was browsing a magazine and came across a double-page spread of this ad, it would also make me stop and read it.

Andy KirkAugust 25th, 2011 at 9:23 am

Thanks for that Andy. You’re spot on with your reference to the images you came across in ‘how things work’ books, those wonderful intricate illustrations, full of colour that made you want to spend ages looking at them. The friction with this MS work is entirely down to the description that goes with it re. “making the data understandable”, but purely as an illustration, perhaps even (stretching the definition to its limits) as an example of data art, it is a really nice execution that engenders pleasure.

Matt RichardsonAugust 25th, 2011 at 11:34 am

I find that’s the same problem I have with some of the infographics I come across. I’m often torn between the beauty/fun element of form and the often very apparent lack of function in terms of getting a point across, especially if that point is communicating data effectively.

I love the comparison to Where’s Wally though, by the way!

Jon PeltierAugust 25th, 2011 at 11:45 am

There is certainly a LOT of information in the cloud graphic, but every piece of it has been hidden in the noise, much like Waldo. To gain any insight, you need to figure out first what is hidden in a section of the graphic, then load the bits of information into short term memory, the reconstruct it in your head, perhaps as a mental chart of some sort. This bypasses the whole reason to use graphs in the first place, because it doesn’t allow us to use our built in cognitive capabilities to interpret anything with our visual cortex. It takes simply too much effort to extract any meaning.

ClementAugust 25th, 2011 at 11:51 am


I actually disagree, I find this visualization interesting for the following reason:

the visual theme of the cloud is applied with many variations to present the results of a survey. Of course bar charts, pie charts, cubes etc.. would have been more effective for the reading, but the attempt was clearly to start first with the visual design of a cloud (and associated air balloons etc.) and then to use it to display data. Of course it does not work perfectly (though the air balloon story is quite straightforward I think), but it is a bold and refreshing attempt to go past conventions.

To make my point differently: datawizzards themselves sometimes acknowledge that some tried and true recipes for infoviz are a bit tired:

In this context, I found this dataviz quite refreshing: trying to include elements of design, and simply being from what we usually see!



ClementAugust 25th, 2011 at 11:53 am

(the last sentence should read: “In this context, I found this dataviz quite refreshing: trying to include elements of design, and simply being *different* from what we usually see!”)

Mike NealeyAugust 25th, 2011 at 2:48 pm

If the only message / information they are intending to get across is that…
Microsoft Cloud Computing is “Accelerating in Asia Pacific”…
…then they succeed.
Because it’s the only clear message within. And (to western eyes, at least), it’s the message posted where our eyes automatically move to first.
What else stands out as I attempt to decipher the information?
The word “cloud” is repeated and repeated and repeated…not to mention the visual clouds that appear.
Again, for me this is emphasizing the link Microsoft are probably intending to make in our minds…Microsoft….Cloud…Accelerating….Asia Pacific….

All in all. Success.

Jan Willem TulpAugust 25th, 2011 at 2:50 pm

I haven’t made up my mind yet on what I think of it. But something that does seem to work is that a real physical world metaphor has been used. This contrasts with many of the big info-graphics Clement is referring to. Here you see a small made-up physical world that you can explore in a similar way to Wally or the How Things Work books. And perhaps the result is a more coherent experience of the visual than several abstract figures like bubbles and squares which is harder to relate to.

I seem to be asking more questions like”what’s happening in this place of this world”, than questions such as “what’s the data at this balloon”.

Andy CotgreaveAugust 25th, 2011 at 2:55 pm

@Jan – it’s a pretty loose “physical world metaphor”, though! A floating lump of rock with some mountains and an enormous drinking straw driven through it.

BTW – this is another classic MS tactic. Does anyone remember when MS tried to get us all to have parties to coincide with the launch of Office x? They released a bunch of awful videos of people partying. End result? Everyone was talking about the Office release date, *particularly* the MS haters. It’s a genius marketing tactic – create something so bad, everyone talks about it. And while they’re talking about it, they’re buzzing about the thing involved. So: terrible infographic or not, we’ve all spent the day talking about MS Cloud Computing in Asia. D’oh. They got us again!

Andy KirkAugust 25th, 2011 at 3:02 pm

Ah yes, the Windows 7 party. What a crazy night that was… ended up modifying a screensaver with some aged middle-class lady I didn’t even know.

Ty FujimuraAugust 25th, 2011 at 3:04 pm

This is precisely why I try to avoid the words “good” and “bad” and their ilk when discussing design, especially data viz. Such descriptions are meaningless without context.

If they actually care about making sure people are aware of these (utterly boring) numbers, then obviously they failed.

But the trainwreck-yness does indeed cement a link between microsoft and the cloud. It’s hard to look away.

Unless they broke the bank for this gem, it’s probably going to product decent ROI. Perhaps even better than an clearer version.

Andy KirkAugust 25th, 2011 at 3:06 pm

You’re right Ty, the language of judgment is important and ‘good’ and ‘bad’ aren’t the correct terms – they do make for a nicely provocative blog post title though!

Thanks for all comments so far, really useful discussion.

[...] Visualising Data » Can a graphic be SO bad it starts to become good? [...]

David FosterSeptember 1st, 2011 at 2:15 pm

I found the first graphic to be quite enjoyable for grazing around. You certainly couldn’t glimpse at it and get the key messages, but it does encourage you to spend time looking at. The second graphic (the air balloon) is frankly, bloody awful (I didn’t say bad ;-)). My first thought was that you were showing us another classic example of a proportion chart that didn’t add up to 100%. They could have made it much clearer and equally ‘pretty’ by having separate hot air balloons for each of the groups.