Four agents, one item: who actually has QC photos

Everyone tells you to check QC photos before you buy. Nobody tells you that the number of photos available depends almost entirely on which agent’s site you happen to look at, and that the gap between the best and worst is more than fifteen times.

So we took one item and looked it up on four agents on the same afternoon.

The item

A single 1688 listing, item id 976641126552, with thirteen SKU variants. Nothing special about it - it was picked because it is ordinary enough that several people have already bought it through several different agents, which is the only condition that matters for this test.

Then the same item id was looked up on four agents’ QC lookup tools, one after another, and the returned photos counted.

The results

Agent	QC photos	Account needed	Photo quality
Sugargoo	96	Yes, logged in	Full size, no watermark
Oopbuy	52	No	Thumbnails by default, full size available
ACBuy	10	No	Full size, grouped by SKU
CSSBuy	6	Yes, logged in	Watermarked

164 photos exist for this one item across the four. If you had checked only CSSBuy, you would have seen six of them and concluded the item was barely documented.

Three things worth knowing

The photos are other people’s orders. This is the part that surprises people. A QC photo is taken by an agent’s warehouse when their customer’s parcel arrives. When you look up an item and see fifty photos, you are looking at fifty strangers’ packages. That is why availability is so uneven: it tracks how many people bought that specific item through that specific agent, not how popular the item is overall.

It also means there is no complete archive anywhere. An item nobody has ordered through any agent has zero QC photos, and no amount of searching changes that.

Two of the four needed no account at all. Oopbuy and ACBuy returned photos to a logged-out visitor. Sugargoo and CSSBuy both required a logged-in session before showing anything. If you are checking an item before you have committed to an agent, that ordering matters - start with the two that do not ask.

Photo counts do not measure agent quality. Sugargoo having 96 photos and CSSBuy having 6 says almost nothing about how well either handles your parcel. It says Sugargoo’s customers happened to buy this item more often. Do not read a QC count as a review.

The variant problem

Of Oopbuy’s 52 photos, they were spread across all thirteen SKUs at roughly four each. That sounds fine until you want a specific colourway - four photos of the variant you actually want is a thin basis for a hundred-euro decision, and two of those four are usually the same angle.

So the useful question is never “how many QC photos does this item have”. It is “how many QC photos does my variant have, and how recent are they”. On an item with thirteen variants, a 52-photo wall can still leave you with nothing.

What we would actually do

1. Start with the agents that do not require a login, to get a read on whether the item is documented at all. If there are zero photos anywhere, that is your answer already.
2. If you have accounts, check the logged-in ones too - in this test they held the largest and the smallest sets, so there is no rule about which is worth checking.
3. Filter to your variant before counting anything.
4. Check the dates before the count.

Method, so you can repeat it

One 1688 item id, four agents’ public QC lookup pages, all within about twenty minutes on 2026-08-10. Counts are the photos each site returned for the item id, across all variants. No agent was contacted, and none of them knew this was being measured. Repeat it on your own item and you will get different absolute numbers - the ratios are what travel.

If you want to try the two that need no account: Oopbuy and ACBuy. Both let you paste an item id without registering.