Test every classification for MECE: no item in two categories, no item in zero categories
For every pair of categories in a classification system, verify that no item can legitimately belong to both (mutual exclusivity test), and verify that no domain item falls outside all categories (collective exhaustiveness test).
Why This Is a Rule
MECE (Mutually Exclusive, Collectively Exhaustive) is the structural integrity test for classification systems. A system that fails either test produces classification confusion: items that belong to multiple categories create double-counting and ambiguous ownership, while items that belong to no category create invisible gaps.
Mutual Exclusivity: for every pair of categories, verify no item legitimately belongs to both. "Is this task a 'bug' or a 'feature request'?" — if the answer is "both," the categories overlap and need redesign. Overlapping categories force classifiers into arbitrary choices that different people resolve differently, destroying classification consistency.
Collective Exhaustiveness: verify that every item in the domain has a home in at least one category. If real items fall outside all categories, the system has a coverage gap. The uncategorized items become invisible — they don't appear in any filtered view, any report, or any analysis.
Both tests must pass for the classification system to be reliable. Most informal classification systems fail at least one.
When This Fires
- Designing a new classification system (priority levels, task types, status flows)
- When classifiers frequently disagree about which category an item belongs to (overlap problem)
- When items keep appearing that "don't fit anywhere" (exhaustiveness problem)
- During any classification system audit or redesign
Common Failure Mode
Testing only one criterion: verifying mutual exclusivity but not exhaustiveness, or vice versa. A system where every item has exactly one home (ME) but 15% of domain items have no home (not CE) is broken. A system where every domain item has a home (CE) but some have two homes (not ME) is also broken. Both criteria must pass.
The Protocol
When designing or auditing a classification system: (1) Mutual Exclusivity test: for every pair of categories, generate a test item that might belong to both. If it legitimately can → the categories overlap. Redesign to eliminate the overlap. (2) Collective Exhaustiveness test: generate 5-10 edge-case items from the domain. Can each one be placed in exactly one category? If any item has no home → add a category or expand an existing one. (3) A system passing both tests is MECE — every item has exactly one home, and every possible item has a home.