Condition Grades Explained: Grade A vs Grade B

A condition grade is a statement about appearance and wear. It is not a statement about function, and collapsing the two is where most disputes start.

On a laptop lot, Grade A means the machines look close to new: minimal cosmetic marking and no damage that draws the eye. Grade B means visible wear consistent with normal corporate use, scuffing on the lid or the palm rest and light marking on the casing, on a machine that is fully functional. Both grades are fully tested, and that testing is a separate exercise from the grade itself.

Keeping the two apart saves arguments later. A Grade B laptop is not a laptop with something wrong with it. It is a laptop somebody used at a desk for three years. Equally, Grade A says nothing about battery health, drive hours or whether the wireless card works, because none of those are visible from the outside.

Read the grade as the cosmetic line and read the test results as everything else. If a supplier gives you one letter and nothing else, ask what the letter covers before you price the lot.

There is no body that defines Grade A, no specification a supplier has to meet before using the term, and no penalty for using it loosely. Every supplier writes its own scale, and the scales genuinely differ.

The practical consequences are worth spelling out. Two quotes for Grade B stock from two suppliers are not comparable on the letter alone. A grade quoted in an email, with no written standard behind it, is an opinion rather than a specification. And a lot that arrives looking worse than you expected is very hard to argue about if nothing was ever written down that it could fail against.

Some suppliers use A, B and C. Some add A+ or use numbers. Some grade cosmetics and function on one scale, which is the least useful arrangement of all, because a machine that looks perfect and a machine that works perfectly are not the same machine.

The fix is simple and it is always available to you: ask for the grading definitions in writing before you agree a price, and hold the delivery against them.

Grading language came out of the laptop and desktop trade, where a cosmetic scale genuinely describes most of what a buyer cares about. Applied to the rest of the categories it covers less and less, until on some items it is nearly meaningless.

The table below is the version of this that is worth keeping. In every row, the useful question is not what letter is on the label but what the letter is standing in for, and what you should ask for instead when the letter cannot carry the weight.

The pattern across the whole table is the same. The more a category's value depends on measured behaviour rather than on appearance, the less a cosmetic grade tells you, and the more the manifest has to carry test data instead. On laptops the grade does most of the work. On drives it does almost none. On a bare chassis, what matters is not condition at all but completeness, because a chassis missing its rails, caddies and fan modules is a different product from one that has them.

On a drive lot the cosmetic grade is close to irrelevant and the test data is the real grade. A drive that looks immaculate can be near failure and a scuffed one can be healthy, because nothing that decides the life left in a drive is visible from the outside.

For spinning disk the numbers that matter are power on hours, reallocated sector count and current pending sectors. Hours say how much service life has been spent. Reallocated sectors say the drive has begun failing and is quietly mapping around the damage. Pending sectors say it is failing now.

For flash the hours are the wrong metric entirely, because a solid state drive wears by writes rather than by time. The equivalent figure is the percentage of rated endurance consumed.

GTR Supply supplies a SMART health report per drive on hard drive lots, so the underlying numbers travel with the manifest and you can apply your own thresholds rather than accepting a letter.

Some categories do not have a middle. A memory module either passes a populated test at its rated speed or it does not, and there is no cosmetic condition that changes the answer. A processor either posts or it does not. Grading these on an A to C scale is dressing a binary result up as a spectrum.

What does vary is the things a grade never covered in the first place. On memory it is the exact part number, because capacity and speed alone do not settle whether a module will work in your board: module type, rank configuration and height each stop a module cold on the wrong platform. On processors it is physical: bent pins on PGA parts, contaminated or damaged contact pads on LGA parts, and whether the part is a low power or OEM specific variant that will not run outside its original platform.

Graphics cards sit between the two. A card can enumerate correctly, report its full specification and still fail under sustained load, which is exactly the condition you bought it for. That is a bench test result, not a letter.

For anything that goes into a rack, the useful scale is not condition, it is completeness. A 2U server missing its rails, its caddies and two of its fan modules is not a Grade B server. It is a server you cannot install until you have bought three things that are harder to find than the server was.

This is why GTR Supply asks sellers to leave rails, bezels and drive caddies fitted rather than stripping them out on the way to the pallet. Rail kits for a discontinued generation can be genuinely difficult to buy at all.

On switches the equivalent list is power supply count, fan modules, rack ears and the firmware level the unit will arrive on. On racks it is doors, side panels, PDUs and the rail hardware inside.

When you are quoting a chassis lot, describe what is present rather than what condition it is in. Present is checkable. Condition is a matter of opinion.

Four things make a grade enforceable rather than decorative, and none of them costs anything to ask for.

Get the definitions in writing before you agree a price, so the words on the quote have a meaning you can point at. Ask what the grade covers and what it excludes, because the exclusions are where the surprises live: batteries, drive hours, missing rails and firmware levels are all routinely outside a cosmetic scale.

Ask for the grade to appear on the manifest line by line rather than as a single description of the whole lot. A pallet described as Grade B is one claim. A manifest with a grade against each line is a hundred checkable ones, and a mixed lot almost always contains more than one grade whatever the headline says.

Then check the delivery against the manifest rather than against your expectation. If what arrives does not match the manifest, that is a clear and checkable failure rather than a disagreement about what was implied, and that is precisely why the manifest is worth insisting on.