In 1990, computer scientist Ben Shneiderman was sharing a Mac with 14 people and wanted to figure out who was hogging space on the hard drive. His solution was to slice the screen horizontally and vertically into boxes that represented proportional usage. Shneiderman called this display a treemap because it gives a complete view of a tree structure, which computer scientists and others use to represent hierarchical information. It offered an overview of the data and also allowed him to zoom and filter, clicking on boxes and drilling down to further detail.

Treemaps have since been used to display information on stocks, news headlines, and much else—Shneiderman gets credit for the concept, if not all the applications. Over the years, he and others have also developed a host of tools for analyzing data in visual form, spotting clusters, outliers, and trends (for one example, see here). These have been a hit among product managers, intelligence analysts, and pharmaceutical companies, Shneiderman says. For instance, a pharmaceutical company might wish to take a library of chemical compounds, filter it according to properties like toxicity, solubility, manufacturing costs, and drug interactions, and end up with a short list of molecules for clinical trials. In these applications, the displays are often brilliant, but for the most part, utility is king.