A Video About Irony

Apple's 2013 holiday commercial:

Think about this one for a sec.

You think they shot & edited that video, about video-making with an iPhone, on an iPhone?

I don't know the answer, just struck by the possible irony.

Charm Factor 5, Scotty

This explanation video from IBM Research is terrific. They turned the MRSA antibiotic-resistant staff strain into an engaging story. Someone did a heroic job of resisting the corporate urge to pile on details and technicalities. I particularly like that they left the puppeteering sticks in plain view.

Note the light-handed branding approach, too.  They don't pound on & on about IBM selling ninja polymers.  It's there, but it's handled factually. (Explanation ≠ commercial!)

The comments on YouTube are interesting.  Some gripe about dumbing down the subject, others say it really engaged them.  The view counter tells the real tale, though - about 29,000 views within two weeks.  Technical, dense explanations don't get that kind of traffic. 

Really nice work, on many levels.



The Technology Alliance Group invited me to talk about information overload last week.  There's a great McLuhan quote - perhaps apocryphal - something like "We don't know who discovered water, but we're pretty sure it wasn't the fish."  Asking just about anyone about information overload is pretty much like asking a fish about water - me included.   However...having thought about this quite a bit, I think it's a bigger issue than we realize.

Ask yourself this, if you're over 30.  Has your reading changed?  I don't mean what you read - that, certainly - but HOW you read. My sense is that I skim & scan much, much more than I used to.  (And "avid reader" doesn't begin to describe my appetite for reading.)

The issue isn't reading and reading habits - it's coping.  Here's the problem, in a visual of course:

The Library of Congress adds about 11,000 volumes per day - total to date, 32 million.

Clive's estimate is the equivalent of 520 million books.

Result:  4,727-to-1. 

What we used to think of as formal knowledge - the stuff that goes in library - is outweighed thousands-fold by the social stuff.  But wait, it gets more interesting.

LOC has 32 million volumes.  Founded in 1815 - 198 years ago.  Multiply by 365 to get 442 volumes per day, on average.  But the current rate is 11,000 per day.  Has the rate of acquisition of formal knowledge gone up?  Sure - dramatically.

I think we're creating a problem that we're not quite culturally ready to 'fess up to (except Ray Kurzweil & his fans.)  Overload?  The word doesn't really capture the issue, does it.  It has a sense of voluntary taking-on-a-bit-much, with a smidge of pride.  We're not "overloaded" - we are incapable of handling this much information in the ways we've handled information for the past 500 years or so.

If you deal with information and knowledge in your profession, you're rightfully proud of both the knowledge you've got, and your skill at adding to it. Saying "I can't keep up" doesn't really square with that very easily. Honestly, though - you're not "keeping up."  I'm not "keeping up."  

The techno-centric response is to point at the knowledge tools we've added - search engines, software, "big data" tools, blah blah blah.  No argument from me, take away my browser and suddenly I'm not half the borg I used to be.  But I don't think my capacity for knowledge acquisition - for adding to what I know and associating it with what I knew - is 4,000 times greater than Grandpa had. It's probably the same;  Jared Diamond says something to that effect in Guns, Germs & Steel as I recall - the guy whose life depends on plants and grass can make a huge number of distinctions, where an idiot like me just sees grass.

Is this a waste of time in the face of the bitflood? Yes - and no. One of the reasons I like working with visual and narrative communication is that it often works when words don't.  It's not a panacea - I don't know about you, but I've started to "read" infographics with the same impatient handling reserved for email.  But it's at least an attempt to reframe communication challenges in light of this big contextual overload issue. 

The rise of "design" strikes me as a response to all this, in a sense.  That's another post.  I have stuff to read ;-)


Data Structure Heatmaps With Excel

I put up a casual post about a data visualization project a while back. Apparently a lot of people are interested in (or fighting with?) Excel for heatmaps, because it keeps getting a LOT of traffic.  Maybe I should capture that traffic and create a heatmap?  Humor.. 

The same project for the client-who-must-not-be-named continues.  They LOVE heat maps.  Every data set, they want a heat map. I've said over and over that it's only relevant when we're making a consistent single comparison, but OK, you wanna heatmap...

I was looking at a very high-level summary, showing pricing across 8 product lines, with 30 varying tiers, for 6 different regions.   (mumble mumble 30 x 6 = 180).  You wanna heatmap, we gotta heatmap (Select Range - Conditional Formatting - Color Scales). I suggest clicking that to see a larger version.

There are really two heatmaps here - the pricing in C:H, and a mirrored % set in J:O.  Just look at the macro-level color pattern - identical, right?

It bugged me a bit, though, because the range is just nuts - from $73.30 to $1,767.24.   Lots of stuff in the yellow middle.  It bugged me because the groups aren't necessarily interchangable, where the tiers might be.

So I tried something a little more elaborate: 

The punchline here, in terms of visual communication, is that this is the same data. If you can, look back and forth at the small versions of these pictures. It's a really different pattern. Wow, those Region 3 & Region 4 guys are good (green). What the heck is going on with Group E in Region 2, and Group F in Region 5. And so on, and so on.

My contention is, there's no way in heck to spot those comparisons by looking at the numbers. It's just not going to happen. Your brain isn't going to hang on to the relationships between 180 different numbers.

OK - quickly - HOW this was done with Conditional Formatting rules. First one - the whole range is a single comparison ("Rule"):

To make the second one - group-by-group - required selecting just the range for a group (e.g. C3:H3 for the first group), and applying conditional formatting for that range. Lather, rinse, repeat.

You can see the list of separate rules here:

Now, I suppose we could have an interesting argument about whether this is a good approach.  The color-coding between Groups is not absolutely consistent. $300 may come up RED in one group, and GREEN in another, and you aren't going to spot the really high prices for the set as a whole, or have a sense for which region or group is relatively low-priced or relatively high-priced. 

Well...yeah, that's why we did both heat maps, but let's leave that aside for a moment. 

The visual-communications question decision/question should be, does one of these color-coding approaches reveal more meaning in the data?  I think the second one does.

It's pretty simple - axiomatic, even.  Because we used additional structure in the data (the Group blocks), we revealed more about the data.  The one-big-block heatmap didn't access that structure, so of course it had nothing to say (visually) about that dimension of the data. 

FWIW I wouldn't have used the color-coding for the different groups (Columns A:B).  It detracts from the meaningful color-coding of the data.  Others' presentation decisions went into that choice. Oh well.

One non-obvious note - spending a little time getting consistent-sized columns, and finessing the gridlines, helps make this a more accurate mechanism for communication. 


PS auto-correct wants "heatmap" to be "heat map."  I don't buy it. Thoughts?

Story Transcends

Did you follow this?

Me too. 

I don't speak Thai either. 

Doesn't matter, does it. 

"A fellow-feeling makes one wondrous kind." 
   David Garrick:  Prologue written for his last performance, June 10, 1776.



Terrific Sankey Diagram from South China Post

I'm becoming more and more a fan of Sankey diagrams.  This one shows how electricity is used in Hong Kong.  It's excellent - click to the page and click again for a legible version.

What makes it terrific? Sounds dumb, but the fact that it's so easy to follow and understand.  Line thickness means something, color means something, and connection/path/destination mean something - and comparing between them is so easy it's unconscious.

Imagine a written report on energy use in Hong Kong.  Think it would fit on a single page, or (more likely) 50-100 pages of prose with a few number tables. 

Imagine a group faced with deciding energy policy, or rates, or infrastructure.  Would having this diagram on the table or the wall help them make better, faster decisions? Would the diagram be regarded as more or less "serious and substantive" as the hypothetical report? Is that an accurate reflection of value, or a cultural bias?   

A client recently rejected a set of Sankey diagrams we'd worked up because they were "too curvy."  Methinks that was more about their culture than the curves, but you can't please all of the people all of the time.  Come to think of it, that proverb would make a good Sankey diagram :-)