MECE is a way of segmenting information into sub-elements that are mutually exclusive and collectively exhaustive. In other words, elements should “exclude” each other, i.e. be distinct and should “exhaust” the relevant field, i.e. contain everything that belongs to it. MECE as a method should be used when you craft an issue tree for your case structure. Doing so will help you avoid dependencies between different branches of the tree and thus sub-problems can be properly isolated.
Mutually Exclusive: aims at reducing complexity by avoiding overlaps. You need to make sure that the possible solutions or groups are not accidentally considered twice. Proving exclusiveness forces you to carefully look at each option, consequently leading you to a much deeper understanding of the issue.
Collectively Exhaustive: aims at ensuring a comprehensive collection without leaving alternatives. Exhaustive means that all possible options have been considered. The trick is to divide a problem into categories with a finite number of general groups. For example, if you need to make “means of transportation” MECE, don't start just collecting “per foot, train, plane, bicyle, bike etc.” Try to find categories, because this practice will prevent you from forgetting important options: wheels vs. no wheels OR air vs. water vs. land.
Another example for a MECE breakdown is dividing the population into subsets of age:
The MECE methodology is famously used at McKinsey. McKinsey interviewers along with other firms will especially look whether you are able to structure information in a MECE format. During a case interview, MECE is also helpful for calculation. Many well-known frameworks, such as Cost-Benefit, 4Cs, Porter’s Five Forces are all designed with the MECE principle in mind.
The example below, from the "Startup vs. Consulting case" on PrepLounge, shows how costs are typically broken down in a MECE way. Subsequently, an issue tree-like format is used to break down the estimation into its basic components for easier approximation of data.