Thursday, April 26, 2012

the color forecast

OK, so far I've been more impressed by this than anyone I've told about it, but I am not going to let that stop me from blogging about it: the color forecast.

What is the color forecast, you might ask? The Pimkie clothing company has installed cameras in Paris, Milan, and Antwerp that record each passersby's clothing, magically extracting the colors and posting them to their website. Not only that, but they provide visualizations that you can use to explore the colors of the day, week, and month. So perhaps forecast is a bit of a misnomer, since we're mostly looking backwards, but let's look past that and explore some of the data.

Here's what clothing colors have looked like recently in Milan:




Normally, I would be offended at a chart that combined as many colors as the ones above (or even a fraction of them). But that doesn't apply when what is being visualized is the color mix itself. Apparently, periwinkle is in right now in Milan. Who knew? Answer: Pimkie, and they are going to use this knowledge to try to get you to buy periwinkle colored clothing:


I still love it.

Tuesday, April 24, 2012

visualizing org structure (or lack thereof)

The new employee handbook from Valve, a gaming company in Bellevue, Washington, has been making the rounds through the HR circle in which I work. It's an amusing read (complete with comics), meant to help new hires understand how to make their place in a company that's trying very hard not to be a company.

They use the following image to help demonstrate the organizational structure:


It is explained with the following text:

Hierarchy is great for maintaining predictability and repeatability. It simplifies planning and makes it easier to control a large group of people from the top down, which is why military organizations rely on it so heavily.

But when you’re an entertainment company that’s spent the last decade going out of its way to recruit the most intelligent, innovative, talented people on Earth, telling them to sit at a desk and do what they’re told obliterates 99 percent of their value. We want innovators, and that means maintaining an environment where they’ll flourish.

That’s why Valve is flat. It’s our shorthand way of saying that we don’t have any management, and nobody “reports to” anybody else. We do have a founder/president, but even he isn’t your manager. This company is yours to steer—toward opportunities and away from risks. You have the power to green-light projects. You have the power to ship products.

Personally, I look at the visuals and accompanying text with a mix of fascination and fear. Given my affinity for structure, this sounds a bit like anarchy to me. It's also unclear to me how this would scale: it may work for a start-up, but will it be possible to maintain as the company grows? In any case, I do think the visuals do a good job supporting the text (and vice versa). I'll leave it at that.

If you're interested in checking out the handbook, you can find it here.

Thursday, April 12, 2012

councours Google de DataViz

Yesterday, I had the pleasure of meeting with the winners of the Councours Google de DataViz: the Google DataViz Contest that ran in France in late 2011. The challenge was to "unleash the power of Twitter and Google data by developing an innovative web application for tracking the French presidential election." Participants were urged to take a new look at the campaign by making an intelligent interface for data vizualization.

Here's a static view of the winning visualization:


I encourage you to view and explore the full interactive version here, created by Stephane Raux, Ronan Quidu, Alexis Jacomy, and Nils Grunwald. In our informal meeting, I told them a bit about the course I teach, we spent time looking at their visual and discussing their work/thought process for generating it, and chatted on data viz in general.

The guys expressed some wonder at winning a data viz contest with a bar chart. I think it's fantastic. One goal they set for themselves early on was to make something that their grandmothers could understand. They had some apprehension when they saw some early postings of other participants' submissions packed with bells and whistles (you can view all the submissions here). We agreed that, while it seems the opposite might be true, making something complicated-looking is actually a lot easier than taking a bunch of data and making it fit together in a way that's easy to understand and explore. This latter approach takes a lot of time and iterations. They commented how useful it was to get feedback from people not close to the project (the graphic designer came in later in the project and while a bit frustrating for him, the others commented on the unintended success of this approach since his fresh perspective meant he could take an outsider's viewpoint which was helpful). 

I really enjoyed the time we spent together and I'm super impressed by their interactive visual that packs in a lot of information and yet is straightforward and easy to explore and interpret. Many of the approaches the guys took and decisions they made were reinforcing of the lessons I teach: keep your audience in mind throughout the design process, use preattentive attributes like color and size to draw attention where you want it, seek feedback from people not involved in the project for a fresh perspective and to direct iterations, and one of my personal golden rules of dataviz: simple beats sexy.

Monday, April 9, 2012

data stories in think quarterly

Today, I arrived back at work after a few days out of the office to find something thrilling at my desk: a hard copy of the latest edition of Think Quarterly. Aside from it simply being a beautiful book, this is the main source of my excitement. It's the first time I've seen my name in print.

The article is short and sweet. But I like to think that this, like the blogging and the conferences and the one-off pieces of feedback I give, is one more step down the path of ridding the world of ineffective graphs.

What step will you take on that path today?