What it is. The third-party economy that breaks if the actor stops. Not the actor’s own revenue, and not a gross “ecosystem” headline — the attributable annual value that others produce because this actor exists and functions.

How we calculate it. Two structural reads, then a dollar estimate:

The published figure is an order-of-magnitude estimate with its reasoning shown, always value-added, never gross. Vendor “ecosystem” studies count total downstream revenue, double-count intermediaries, and skip the counterfactual — the standard critique of App-Store-style numbers. We deflate to what’s attributable, and say what the dependents would earn without the actor when we can.

The canonical example. Arm books ~$5B of revenue; its instruction set is in ~99% of smartphones, and Apple alone earns ~$400B a year on Arm-based silicon. Weight tells you Arm is small. Gravity tells you the world can’t turn it off. The gap between the two axes is the whole reason both exist.

Confidence, honestly. Gravity is the lowest-confidence axis by design — there is no filed document for “who breaks if we stop.” Every gravity claim shows its rationale, and most carry a low-confidence flag. Treat them as directional coordinates, not audited figures.

Prior art. The Basel Committee’s G-SIB substitutability score is the working, published version of this question for banks; forward linkage/upstreamness is the input-output economics version; Gabaix’s granularity is the macro-volatility version; weaponized interdependence (Farrell & Newman) is the geopolitical version — dependency that is also a chokepoint. We borrow the constructs and keep the arithmetic rough.