Visually stunning work from Christian Stoll, it really captures the idea of scale and the result of scaling horizontally. I wonder if the architects imagined that the escalator machinery had to be as wide it appears to set the escalator stairs as narrow as they look…

I recently finished reading Dan Roam’s “The Back of the Napkin,” which is a great introductory and reference book on communicating with pictures.  I also spotted Seth Godin’s blog post commenting on this NYTimes post and its validity.  And last but not least, one of my favorite examples of multiple dimensions of data plotted on a single illustration is explained by Edward Tufte.  I mention all of these examples because it’s easy to make a bad picture, but not too hard to make a really good one.

My career is filled with all sorts of illustrations to prove a point. Whether it was as pedestrian as a pie chart, or a project timeline, or a business roadmap, illustrations allow someone to demonstrate a few things:

• That you’ve researched the problem at hand, and know what you’re talking about.
• That you can show variations of the issue (by drawing on top of the picture).
• That your analysis is sound.

There’s an inherent danger to pictures though, which is that sometimes they lend too much credence (as Godin points out above) to a weak analysis.  Godin points out that

“reviews never reflect the product, they reflect the passion people have for the product. As Jeff Bezos has pointed out again and again, most great products get 5 star and 1 star reviews. That makes sense… why would you be passionate enough about something that’s sort of ‘meh’ to bother writing a three star review?”

The point for any viewer of the illustration is to be skeptical of the data that underpins the picture.  Note that the NYTimes post relies on shaky data, whereas the Minard illustration is based on mostly secondhand (but reliable) accounts.  Both require at least two passes across the pictures shown, and the biggest difference is that the NYTimes post discredits itself, whereas Minard’s builds on itself.

Conclusion: Check to see if your pretty picture relies on not-so-pretty data.  If it does, don’t be lazy; find better data (i.e. garbage in, garbage out).