Definition
A method of low cost usability engineering proposed by Jakob Nielsen that asserts that usability that is performed frequently with fewer participants, and using scenario based observation yields results that are nearly as good, and with better ROI when compared with costlier traditional usability testing.
Ergosoft Laboratories summarizes the method:
The primary characteristics of discount usability engineering are small sample sizes (i.e., fewer participants), frequent iterations of these small tests, and reliance on direct observations rather than on statistically established findings. Discount usability engineering is best suited to projects that have access to an in-house staff of usability specialists. A prerequisite for discount usability engineering is the capability to quickly incorporate the results of each small study into a subsequent iteration of the prototype which will be tested again. Likewise, it's necessary to be able to define test tasks quickly and perform the testing in a short timeframe. These short turn-around cycles are easiest to perform when all the developers and experimenters involved work for the same company and have an established pattern of working together.