Often all you need is the green pencil. Or maybe the red, blue and green pencils. You may need a subset of all the other pencils, and there may be a lot of pencils. We can sample subsets of text documents in a structured way. We can identify groups based on analytics or word searches or both. We can derive samples of groups, extracted at random or by criteria, and export them for further analysis. Segmentation is critical to understanding any data set, especially text.