The secondary device market runs on three letters: A, B, and C. Every listing, every wholesale price schedule, every carrier trade-in offer references these grades as if they're a shared standard. They're not. Grade A from one operation means something different from Grade A at the next one, and the spread between interpretations shows up in pricing disputes, buyer returns, and lot-level renegotiations.
Understanding what condition grades actually measure, how image-based assessment maps to those measurements, and how grade uncertainty translates into price uncertainty is worth spending time on for anyone operating in device intake or resale. This is that article.
What a Grade Actually Captures
Condition grades for refurbished handsets combine two fundamentally different assessments: cosmetic condition and functional condition. Most dispute and inconsistency in grading comes from the cosmetic side, so it's worth separating the two.
Functional condition is relatively binary. A device either boots, runs through its diagnostic checks, has a working screen, speakers, microphone, cameras, and ports, or it doesn't. Functional assessment is done with diagnostic software, not by eye, and it either passes or fails each check. Where functional grading becomes nuanced is in borderline cases: a battery at 78% health that passes a diagnostic threshold but will degrade to below threshold quickly, a touch screen that's 98% responsive but has a dead zone near one edge, a speaker that measures within spec but audibly distorts under load. These edge cases require judgment, and consistent judgment requires defined thresholds, not general rubrics.
Cosmetic condition is where the real grade complexity lives. A cosmetic grade summarizes visible wear across the screen, the chassis, and the back panel into a single letter. That compression loses information, but the market prices on it because a single letter is tractable. The question is how much compression happens before the letter loses its signal value.
The Cosmetic Dimension Breakdown
The cosmetic characteristics that drive grade differences in buyer acceptance break down across a few distinct dimensions. Screen condition matters most: scratches, cracks, dead pixels, and touch sensitivity issues are the highest-priority cosmetic factors in every buyer acceptance framework we've seen. Buyers who will accept back panel wear or minor chassis scuffs without a price concession will reject the same device if the screen has visible wear under ambient light.
Back panel and chassis condition matters second, with significant variation by device category. For premium-tier devices where the back panel is a key part of the industrial design, back glass condition carries more weight. For mid-range devices, chassis scuffs are generally lower priority in buyer frameworks.
Port and button condition matters third: physical damage to the charging port, speaker grille, or volume buttons affects functional concerns as much as cosmetic ones, and buyers treat it differently from pure cosmetic wear.
Grade A typically captures: screen with no visible scratches under ambient light, chassis and back panel with no visible scratches or chips, all ports and buttons functional and cosmetically intact. Some operations add a requirement for original factory settings and original accessories in packaging, which affects where grade A sits in their pricing model but isn't a universal standard.
Grade B captures: minor screen wear visible only under direct raking light or at specific angles, light chassis wear, back panel scuffs not affecting structure. The key word is "minor," and "minor" is the locus of all grading disagreement in the industry.
Grade C captures: visible screen wear under ambient light, meaningful chassis wear, possible non-structural cracks on back panel. Grade C devices are still functional; the condition description is cosmetic.
How Image Data Maps to These Criteria
Image-based grading uses structured photo captures taken under controlled or standardized lighting conditions, and applies assessment criteria to those images to determine which category each visual characteristic falls into. The advantage of image-based assessment over direct visual inspection isn't just speed: it's that the image creates a fixed record of the condition at assessment time, with consistent lighting that reveals certain types of wear (particularly micro-scratches on screens and back panels) more reliably than typical retail lighting environments.
Raking light, where a light source at a low angle across the device surface, is the most revealing condition for screen micro-scratches. Images captured with a standardized lighting setup that includes a raking-light component will reveal wear that a direct overhead light inspection misses. This means that image-based grading under consistent lighting will systematically detect more cosmetic wear than a manual inspection under retail lighting conditions, and will therefore tend toward lower grades on borderline devices. This is not a flaw in image-based grading. It's a feature: the grade reflects what buyers will actually see when they inspect the device, because buyers who care about condition will check under raking light.
For a buyback desk, the implication is that switching to image-based grading may initially produce more Grade B grades on devices that would have been manually graded as Grade A. Before interpreting this as a system problem, it's worth verifying against what actually happens when those devices reach buyers: do they accept them as Grade A, or do they flag the screen wear? In our own tests with a few hundred handsets, the image-based grades tracked buyer acceptance thresholds more closely than the manual grades they replaced.
The Resale Value Equation
Resale value for a given device model is not one number; it's a range defined by grade and market. Grade A devices in a given model command a meaningful premium over Grade B, and Grade B commands a meaningful premium over Grade C. The exact spread depends on model, market, and supply-demand conditions, but the structure is consistent: each grade step represents a significant portion of device value.
For common mid-range handsets in the Southeast Asian secondary market, rough observed ranges (not definitive benchmarks, based on internal data from a limited sample) suggest Grade A to Grade B gaps in the 10-20% range and Grade B to Grade C gaps in the 15-30% range. These spreads compress when supply of a specific model is tight and expand when supply exceeds demand. The grade matters most for high-supply models where buyers have selection and can afford to be choosy.
The practical consequence: the margin difference between accurately grading a device as B versus over-grading it as A is not just the immediate acquisition price. It's also the downstream disposition: a Grade A with hidden Grade B cosmetic condition generates returns and buyer relationship friction. The real cost of grading inflation is in the dispute chain, not just the intake price.
Making the Grade Meaningful
A condition grade is a commitment about what a buyer is receiving. When the criteria are precise, the assessment methodology is consistent, and the documentation captures what was actually evaluated, the grade carries real information. It's the compact encoding of a structured inspection result, and it's the basis for every pricing decision in the resale chain from intake to end buyer.
When criteria are vague, assessment is inconsistent, and documentation is just a letter on a form, the grade is noise. It occupies the same position in the transaction but carries none of the information function. The transactions still happen, but disputes, returns, and pricing adjustments replace the signal work that grading should be doing.
The goal isn't to have a perfect grading system. It's to have one consistent enough that Grade B means the same thing today as it did last month and at this location as at your other locations. That consistency is what makes the grade useful as a pricing basis, a contract reference, and a buyer communication. Without it, A, B, and C are just starting positions for a negotiation.