Monthly Archives: February 2010

Unemployment insight dashboard for Jan 2010 shows 41% of all unemployed, 6.3MM people, out of work more than 6 months.

My dashboard of unemployment in the U.S. is updated with data from January 2010.

I’ve added sparklines to the Demographic section in the middle of the dashboard. Now, rather than just seeing where unemployment stands this month for a particular demographic segment, you can see where it’s been over the last 12 months. The sparklines on the left side of the demographic section all represent the unemployment rate over the last 12 months. The sparklines on the right show the percentage of total unemployed each segment represents over the last 12 months. For example, a value of 25% in the “% of total unemployed” column for the “White Women” segment means that White Women make up 25% of all unemployed.

It’s interesting to see the huge drop, 9.5%, in underemployment this month. Check out the small chart in the upper right hand side of the dashboard. We went from 9.1MM underemployed in Dec to 8.2MM in Jan 2010. That’s the first big drop in underemployment in at least 12 months.

The long-term unemployed population, however, continues to grow. Another 200,000 Americans added to the pile of 6.1MM Americans who’ve been jobless for more than 6 months. Those long-term unemployed are equal to the entire population of Tennessee. The NYTimes had an indepth story on the long term unemployed, they call them “The New Poor.” You can find that story here.

New data from the Bureau of Labor Statistics will be released this Friday, the 5th, so subscribe to my blog and you’ll get an email notifying you when the revised dashboard is complete.

As always, your feedback is welcome.

Click on the image to enlarge.

If you’d like a beautiful 11 X 17, crystal clear pdf of my dashboard, click here.

BTW, this dashboard was done using Excel 2007.

Dashboard of Joblessness in the U.S.-Jan 2010

Thoughts on “7 Rules for Dashboard Design” post on Dashboard Insight

This post addresses a post on Dashboard Insight’s site titled, “7 Small Business Dashboard Design Dos and Dont’s”

Hi Stacey,

I read your post with great interest, after all, who wouldn’t like to know the 7 rules of dashboard design? As soon as I got to the 19th word of your post, “colorful” I knew that we’d have some interesting differences in our viewpoints.

So let’s start with how you define a dashboard. I agree with most of your definition, especially the part about face-smacking insights. Certainly a useful dashboard should provide the reader with insights. But does a dashboard need to have “gorgeous colorful graphs?” I think not. In fact, if the dashboard designer uses too many colors in their graphs, they can kiss those face-smacking insights goodbye. I think it would be better to say that an insightful dashboard needs color use to be, restrained. The designer needs to be sparing with his/her use of color so that when color is used, the reader’s eyes are immediately drawn to the thing that the designer is emphasizing. Use color everywhere and it becomes meaningless.

I agree with your rule #1 (start with a few key business metrics, don’t waste time collecting everything), and whole-heartedly agree with your rule #2 (use basic tools that you already have). In fact, I’d recommend that most people, beginners and experts alike, use the combination of Excel with XLCubed. XLCubed gives the designer the ability to create very small and crisp sparklines, microbarcharts, bulletcharts, and other useful graphics.

Rule #3 could be summed up as two rules. First, use simple line charts. And second, don’t try to compare things (this month to last month) on your dashboard, it’s never a good idea.

I’m all for simple charts, after all, most complicated charts are a result lazy design (exception is Napeloeon’s march). A good dashboard designer takes the time to ensure that his/her charts are almost instantly insightful. Complex, hard to read charts are almost never instantly insightful.
The second part of your rule #3, that is, don’t make two point comparisons, really surprised me. The most insightful dashboards I’ve seen are ones where the reader can instantly see if a critical measure is above or below last month, or last year, or if a measure is over or under forecast. For that, you absolutely have to compare two points. Further, check out Stephen Few’s bullet charts. They’re all about comparing two things, actual to forecast, test to control, this to that, and that to this. In fact, one rule all dashboard designers might want to follow is to ask themselves this question when including a measure on their dashboard, “Compared to what?” I got that one from Stephen Few.

Rule #4 about Pareto charts, I’m just not a big fan. Yes, the 80/20 rule is an important one to know about, but the actual Pareto chart violates your first rule of rule #3, it ain’t that simple. Dual Axes with a line and a bar on the same graph just isn’t all that intuitive. I agree with your second rule of rule #4, no pie charts, although some info viz bloggers are pushing back against that one, like Jorge Camoes here.

Rule #5, don’t go all Picaso on it. Agree. I agree too when you say ‘limit your creativity to use of colours,’ so long as you mean that colours should be used sparingly and to call out the most important things to monitor on the dashboard.

Rule #6, Monitor your dashboard weekly. Agree 100%. Monitor that dashboard. After all, if you don’t check it out to see how things are doing, what good is it?

Rule #7, Get help from someone who’s ok with Excel and charting, not from someone who’s selling you the software. For most BI vendors out there, I agree with your assessment.

I applaud your efforts to put down rules for small business dashboard designers to follow. And your points about well-designed dashboards helping business owners monitor their businesses is right on point.

Lastly, there’s a small and passionate community of dashboard designers on the web. You can find some great advice from Stephen Few’s website. He’s also published 4 excellent books about information visualization, one focusing exclusively on dashboard design.

Thanks,

John C. Munoz