All product is not created equal. Some units are more likely to fail in service than others. Thus, in reliability evaluations, you need to identify subpopulations with different failure susceptibility. This is accomplished through segmentation—a divide-and-conquer strategy that breaks down the product population into meaningful subpopulations so you can conduct separate analyses on each and then act on the resulting information.