Last weekend's New York Times Magazine ran an article entitled "From 0 to 60 to World Domination" about Toyota's remarkable success. Some of the reasons cited were its focus on long-term strategy (such as fuel economy and the youth market) enabling it to be market leaders when the moment was ripe, the just-in-time production system freeing up loads of cash for marketing and R&D, and kaizen, or continuous improvement.

One of the methods for achieving kaizen was obsessive field research, and the article detailed the personas used in developing and then marketing Toyota's first full-size truck, the Tundra:

1) fishers and outdoorsmen; 2) home-improvement types; 3) Nascar fans; 4) motorcycle enthusiasts; and 5) country-music lovers.

What's strange about this picture? None of these categories are necessarily truck drivers, or even comment directly on driving needs. (In the vein of, say, road trippers, cargo haulers, errands-around-towners.) They're all orthogonal to the product.

When you focus specifically on usage of the product, you end up with feature-driven needs like "rack for sheetrock", "great handling" or "good sound system". Go one step backward to the contexts surrounding their usage, and you'll find needs like "big gear knob" because, for example, construction workers often leave their work gloves on all day.