COCOMO stands for

Algorithmic cost-estimation models in software engineering predict the effort needed to build a system by feeding measurable inputs — most often an estimate of the software's size, such as in thousands of lines of code (KLOC) — into a mathematical formula. Depending on the model's variant, the raw result may then be refined further using cost-driver adjustments that capture factors like team experience, tool support, and project complexity.

Barry Boehm introduced exactly this kind of model in 1981, publishing it in his book “Software Engineering Economics.” He named it using the acronym built from the words Constructive, Cost, and Model, signalling that the model constructs (builds up) an effort figure step by step — starting from a size estimate and a mode-based coefficient pair (a, b), refined further by cost-driver adjustments in its Intermediate and Detailed variants — rather than treating the estimate as something already spent or predetermined.

• “Consumed cost model” does not fit: nothing in the model's design tracks resources already used up — it estimates effort in advance of development, the opposite of a “consumed” framing.

• “Common control model” does not fit: the model's letters do not correspond to “common” or “control,” and its purpose is a size-to-effort calculation, not coordination or governance.

• “Composition cost model” does not fit: “composition” implies pricing out the assembly of existing components, a different estimation task from the size-driven formula this model actually uses.

So the acronym expands to Constructive Cost Model — the algorithmic effort-estimation model developed by Barry Boehm, built around the formula Effort = a × (Size)b, where a and b are constants set by the project's mode — organic, semi-detached, or embedded. COCOMO itself comes in three variants of increasing detail: Basic, Intermediate, and Detailed COCOMO (later extended as COCOMO II).