I've got a typical impedance mismatch where the data is stored in a normalized fashion:
Client | Year | Value -------------------------- Foo 2007 9000000 Foo 2008 10000000 Bar 2008 20000000 Foo 2009 12000000 Bar 2009 15000000
But the users consume, visualize and comprehend it in a denormalized fashion:
Client | 2007 Value | 2008 Value | 2009 Value --------------------------------------------- Foo 9000000 10000000 12000000 Bar - 20000000 15000000
Which is all well and good—I've got a variety of solutions in place for pivoting and unpivoting the data between rendering and persistence. However, the in the real-world case the denormalized rendering has dozens of columns, which becomes unwieldy to use and to get ones head around.
Keeping in mind that the users currently grok the data in its denormalized fashion, and that I don't want to mess with their mental map to try to change that, what are some strategies for better presentation of wide tabular data? I know it's not a new problem but I don't know where to look for good insights.
I'm not looking for answers specific to any one presentation environment. Whether in a web browser or a spreadsheet, grasping a wide presentation of data is simply a mentally difficult task and I'm looking for general strategies or presentational shifts—or pointers to resources—to mitigate that difficulty.
A picture can go a long way. A graph of the data can provide people which what they are usually looking for - patterns and abnormalities. In the case of your denormalized data table, I'd suggest adding a spark chart on each row that shows all the values. That way looking down the list one can see how the different values changed.
Here are some examples of people using inline charts in their data to provide visualizations of hard numbers:
http://s.wsj.net/public/resources/images/P1-AL682_OILFIE_20080521185242.gif
http://mjsbigblog.com/wp-content/uploads/2008/02/billboardupdate020908.png
http://www.economist.com/images/20041009/CUS756.gif
Edward Tufte has some exceptional guidelines on visualizing data, check out the sparklines for the Wall Street stocks example on his wikipedia page.
