Grading: Four Dimensions, One Letter

A grade is a contract. When a buyer pays for "Grade B," they have an expectation. When the seller delivers something different, the dispute opens. Grading exists to make the contract precise — not to be artistic.

The four sub-grades

Functional (F0–F3): does it work? F0 = fully functional, F3 = does not power on. Battery-related issues live in their own grade because a working laptop with a dead battery is not the same as a laptop that won’t boot.

Cosmetic (C0–C4): how does it look? C0 = like new, C4 = severe damage. The cosmetic grade is the most subjective of the four, which is why the defect-codes-and-zones structure exists — to anchor it to specific observations.

Battery (B0–B3): B0 = >85% design capacity, B3 = <50% or won’t hold charge. Computed from the battery-health check during testing. For non-battery devices (desktops, servers), this dimension is N/A.

Data (D0–D3): data-erasure status. D0 = certified erasure complete, D1 = erasure in progress, D2 = pending, D3 = erasure failed (returns to the queue). This is the dimension that gates resale: D0 is the only sellable status.

The overall A–R grade

The overall grade is what buyers see on listings. A–G covers the saleable spectrum (A is best); H–N is the recycling spectrum; P–R is the parts-or-recycling end. The mapping from sub-grades to the overall letter is computed by a tenant-configurable rule set: by default, cosmetic 30% + functional 50% + battery 20%, with severity-4 defects forcing a floor of D regardless of the math.

Why the rules live in settings

Different markets have different definitions. A broker selling to refurbishers is stricter on functional than a recycler buying for parts. The tenant settings page (/settings/core/grading-rules) lets each org tune the weights and the floor rules. The rule version is stamped on every graded asset, so if the rules change next quarter, last quarter’s grades don’t silently shift.