Your Grade A Could Be Someone Else's Grade B

Your Grade A could be someone else's Grade B.

Anyone who sells refurbished laptops has lived this. A lot leaves your floor as Grade B. The buyer opens a few boxes, calls half of them C, and asks for a credit. Nobody lied. You just meant different things by "B".

That's two problems, not one. There's no shared definition of a grade between companies. And inside one company, people apply the definition they do have differently.

---

1. There's no common scale for laptops

The phone market tried to solve this years ago. CTIA, the US wireless industry association, published grading scales for pre-owned wireless devices in 2019 and has revised them since. In May 2024 it aligned them with SERI's R2 Equipment Categorization. On June 10, 2026 it released Version 5.1, which adds guidance for "mapping wholesale market device condition grading criteria into direct-to-consumer relatable terms", plus requirements for battery health and genuine-parts declarations. CTIA says the scales let buyers and sellers "utilize a common grading definition throughout the industry."

That covers smartphones and tablets, and CTIA also publishes criteria for Chromebooks. For the Windows laptops and Macs that fill most ITAD floors, nothing comparable is shared. Each buyer, marketplace and refurbisher writes its own.

And most of those definitions are written in words, not measurements. eBay's refurbished program, for example, defines its tiers like this:

Those are reasonable descriptions for a shopper. But "minimal" and "moderate" can't be measured at a bench. One technician's minimal is another's moderate.

2. Even one floor grades the same scratch two ways

Suppose you do write your own standard. Grades still drift, because looking at a device and judging it is hard to do the same way twice.

The best evidence comes from outside ITAD. In a 2015 study published in Human Factors, Judi See of Sandia National Laboratories had 82 trained inspectors examine 140 precision parts for eight kinds of defect. They correctly rejected 85% of the defective parts. They also rejected 35% of the parts that were fine. These were trained professionals, and they still missed some defects and flagged good parts.

A laptop lid is easier than a machined part. But the lesson carries over: the looser the definition, the more room each person has to read it their own way. "Light scratches" leaves a lot of room. "No scratch longer than 2 cm on the lid" leaves much less.

3. What inconsistent grades cost

Inconsistent grading doesn't show up as one big failure. It leaks out in small ways:

• Credits and disputes. A buyer who regrades your lot down will ask for money back, or pay less next time.
• Re-grading. Someone opens the boxes and grades the lot again, which is labour you've already paid for once.
• Pricing you can't defend. If two benches would grade the same unit differently, your grade mix is partly noise.
• Trust. A buyer who's been burned grades every lot from you themselves, and prices in the risk.

None of this needs a new scale for the whole industry to fix. It needs your scale to be clear, and applied the same way every time.

4. How to make a grade mean one thing

Here's what consistent grading looks like, whatever tools you use:

1. Write it per part. Lid, bezel, palm rest, keyboard, screen, base. A grade for "the device" hides where the wear is.
2. Make each defect concrete. A scratch you can feel with a fingernail, a crack, a missing key, a dead pixel. Where you can, use a measurement.
3. Give each defect a ceiling. "A cracked screen can't be better than C." The device's grade is then the worst part's grade, and nobody has to weigh it up in their head.
4. Ask the same questions, in the same order, at every bench. If every technician answers the same questions, the grade comes from the answers, not the person.
5. Let functional results limit the grade. A device with a failed keyboard test shouldn't leave as Grade A, however clean it looks.
6. Keep one standard per buyer when buyers differ. If one client grades stricter, grade their lots against their standard, not yours with a note.
7. Send the definition with the grades. A buyer who can read what your "B" means can check it, instead of guessing.
8. Record who graded what. When a dispute comes, you want the grade, the person, the time and the standard it was graded against.

5. How reCore does it

We built reCore's grading around those eight points. Here's what it does today, checked against the code.

You write your standard once, in the dashboard. Under Grading & Standards, a standard lists the parts, the defects each part can have, and a grade ceiling for each defect, such as "Max B". If a part matters to you that isn't in the list, you add it and write its question. The grade itself is the worst part's grade.

Every station asks that standard's questions. At the bench, the station shows which standard it's grading against and walks the technician through each part. One key per answer. From the answers and the ceilings, it works out the suggested grade. Same answers, same suggested grade, on every station.

Tests set a limit. A failed test caps the grade: a failed display, keyboard or camera test, for example, caps it at the bottom of the scale. If a technician overrides the grade above what the tests allow, the station warns them.

Different clients, different standards. You can give a batch its own standard, or set a default standard for a client, so a stricter client's lots are graded against their rules automatically.

The buyer gets the definition. reCore's stock list, a spreadsheet you can send with a lot, includes each unit's grade, its recorded defects by part, battery health and sanitization status, and a Grade Guide sheet that gives each grade's meaning as your standard defines it.

There's a history. Every grade is saved as a new record with who graded it, when, and which standard it was graded against, and the device page shows the history.

Conclusion

Until laptops get a shared scale like phones have, every "Grade B" in this industry is somebody's own definition. That's fine, as long as it's written down, applied the same way at every bench, and handed to the buyer with the lot. The disputes don't come from grades being different. They come from nobody being able to say what a grade meant.

See how reCore handles grading, or how workflows let each line run only the steps it needs.

---

Sources

• CTIA, CTIA Publishes Wireless Device Grading Standard in Support of Greater Transparency for Pre-Owned Devices, June 10, 2026 (Version 5.1).
• CTIA Certification, Grading Scales (smartphone and tablet criteria V5.1; Chromebook criteria V1.0).
• CTIA Certification, CTIA Updates Wireless Device Grading Standard to Help Extend the Lifecycle of Pre-Owned Wireless Devices, May 9, 2024 (alignment with SERI's R2 Equipment Categorization).
• J. E. See, Visual Inspection Reliability for Precision Manufactured Parts, Human Factors 57(8), 1427-1442, 2015.
• eBay, Item condition ID values (refurbished condition definitions).