── ── Decision-making

MECE

MECE (Mutually Exclusive, Collectively Exhaustive) is a principle for structuring a problem so its parts do not overlap and together cover the whole. Each item belongs to exactly one category, and no relevant item is left out. It produces clean, gap-free decomposition that prevents double-counting and blind spots in analysis.

How it works

You take a messy question and split it into categories chosen so nothing falls into two buckets at once (mutually exclusive) and nothing is missing (collectively exhaustive). Common cuts include by stage, by segment, by cause, or by simple numeric ranges that provably tile the whole space.

Test each cut twice: could any item land in two buckets? Then it's not exclusive. Could a real item land in no bucket? Then it's not exhaustive. Tighten the categories until both checks pass, then analyze each branch independently.

When to use it

  • Sizing a market or breaking revenue into drivers you can act on separately
  • Diagnosing why a metric dropped without missing a possible cause
  • Structuring a strategy deck or board memo so the logic is airtight
  • Dividing work across a team so responsibilities don't overlap or fall through

When not to use it

When the problem is genuinely interconnected and forcing clean separation hides the feedback loops that actually drive it.

Worked example

Sizing a market the clean way

To estimate demand for an electric toothbrush, you can segment the population by mutually exclusive age bands (0–17, 18–34, 35–54, 55+) that together cover everyone. Because the bands neither overlap nor leave anyone out, you can size each independently and sum them without double-counting. The same population cut by 'young, urban, health-conscious' would overlap and leave gaps.

Why it matters for founders

When you tell investors why growth stalled, a sloppy answer that double-counts causes or misses one entirely costs you credibility and the right fix. MECE forces you to name every driver once and only once before you commit resources. deciqAI's agents structure a problem this way before acting, so the plan covers the whole board instead of the part that was top of mind.

Install this skill (free, MIT)

$npx skills add deciqAI/knowledge-skills
View MECE source on GitHub →

FAQ

Is MECE just making a list?

No. A list can have overlaps and gaps. MECE imposes two strict tests: every item fits exactly one category, and the categories together cover everything relevant.

Does a breakdown have to be perfectly MECE to be useful?

Perfection is a discipline, not a dogma. The value is in catching overlaps that inflate your numbers and gaps that hide a cause; even an approximately MECE cut surfaces both.

Start free. Pay when it pays off.

These skills are open source. deciqAI is the operator team that runs them — autonomously, on your company.

Start free