Why Your DHgate Spreadsheet Link Opens Something Else

Which shell you landed on

A spreadsheet row is a pointer, not a product page. The row holds a short code, a thumbnail and a price band; the page it opens was written by whoever listed the item, and that page can be repointed, relisted or replaced without the row changing at all. The desk reads the same report in different words most weeks: the link resolved, the price looked close, and the item on screen was not the item the row described. Naming the shape of the page comes before any judgement about the seller, and it is the step most readers skip.

Four shapes cover almost every case grouped from public threads so far. One is a single-item page with its own photographs and its own title. Another is an accessory page standing in front of the main item. A third is an unbranded page carrying generic wording. The fourth is a bundle page that mixes the item with unrelated filler. Each shape fails differently and each needs a different recovery move, which is why classification has to happen before the photographs are read for quality.

Classify from the page, never from the row. The row supplies expectations; the page supplies evidence. Read the page title first, then the seller's own gallery, then the specification block, and write down where those three disagree. The disagreement is the finding, not a nuisance on the way to one. Keep the note short enough to be usable later: page title as written, item named, item photographed, and the date the page was read. Two rows that look identical in a list can resolve to two different shapes, so the note has to describe the page that was actually opened rather than the row that advertised it.

The distinction matters because the four shapes have different escape routes. A decoy accessory can often be exchanged for the real item through the option list, if the option exists and the seller agrees. A stand-in cannot be exchanged, because there is nothing on the page to exchange it for. A bundle can be clarified with a packing list. A page that merely looked wrong is usually a stale row rather than a deceptive listing, and that case needs no message at all.

Accessory decoy branch

The accessory branch is the cheapest shape to build, which is why it dominates the recorded sample. A seller lists a phone case, a dust bag, a keyring or a pair of laces at a low figure, attaches photographs of the finished item, and lets the search result carry the rest. The title is honest in a narrow sense: the product named in the listing headline really is the case, and the case is what ships.

Detect it by reading the title and the gallery at the same moment. Where the title names a part, a component or an accessory while the gallery shows the whole item, the page is a decoy. Check the option selector as well, because decoy pages often park the real item in the options at several times the visible price; that keeps a low number in the row and a different purchase behind the link.

Recovery starts from the row rather than the page. Copy the item description out of the row, search that phrase together with the storefront name, and compare any result against the original wording. Where no equivalent page exists, the working conclusion is that the row pointed at a placeholder and the item it promised may never have been listed at that price at all. Where a seller confirms that the anchor item was never available at the row price, that answer is worth keeping, because it turns a suspicion into a documented correction.

Price is the cheapest clue on a decoy page and the easiest one to misread. A figure far below the band its category normally occupies is a signal, but so is an ordinary figure attached to a title that names a part. Read the price of the anchor item inside the option list rather than the price in the row, then compare that figure with the band. The gap between the two numbers is the size of the substitution.

Unbranded stand-in branch

The unbranded stand-in keeps the outline of the original listing and removes the identity. Photographs still match the item in silhouette; the title drops the model name; the specification block fills with phrases such as premium quality, fashion new and hot sale. Such a page is not pretending to be a different product. It is declining to name one, which leaves nothing for a later comparison to attach to.

This branch is the hardest of the four to call, because generic wording is normal on listings that carry no brand claim. The usable test is whether the page can be matched back to the row on a second identifier. Batch codes, colourway names, size runs, box labels and model numbers all belong to the identity of an item, and a page that will not state any of them cannot be checked against the row that sent you there.

Once a stand-in is confirmed, treat the price band as unverified. A generic page can be swapped to a cheaper supply at any point, since nothing printed on it ties the seller to a fixed specification. The follow-up action is a written question naming the exact model, colourway and size, with the reply kept before any money moves and the question phrased so a vague answer is visible as vague.

Generic pages also defeat the reverse test, since they cannot be told apart from a genuinely unbranded product sold honestly at a low price. That is the method working rather than failing. Where a page states nothing, the accurate classification is that nothing is known, and the decision has to rest on price and on the seller's written answers. Filing that as unknown is more useful than forcing a verdict the page cannot support.

See current listings

Multi-item bundle branch

A bundle page lists several unrelated products under one price, and the row pointing at it usually shows the photograph of the most attractive one. Socks, small tools, stickers and phone accessories appear as filler; the anchor item in the gallery is what the buyer believed they were paying for. Bundles are legal, common and occasionally generous, so the shape alone proves nothing about intent.

Read a bundle from the option list downwards. Count how many distinct products the page names, then note the price attached to each option. Where the visible figure belongs to the cheapest option and the anchor item sits several times higher, the row and the page are describing two different purchases. The distance between those two numbers is the size of the mismatch, and it belongs in your note.

The bundle branch carries a risk the other three do not: quantity. Some bundles deliver a set of the anchor item, some deliver one item plus filler, and the page rarely states which. Ask for the packing list in writing, name the unit count you expect, and keep the answer. A packing list also settles a later dispute faster than a photograph of a parcel does.

Bundle pages distort the price history that a row carries. A row showing a low figure because the cheapest option is filler will keep showing that figure after the option list changes, and a reader returning later compares the new number against a baseline that never described the anchor item. Recording which option was selected, rather than only the figure seen, keeps that comparison honest across visits.

Cost of assuming the photo matches

The cost of landing on the wrong shell is not the price of the item. It is the round of messages, the waiting, and the evidence nobody kept. A buyer who assumes the gallery belongs to the page title has nothing to file, because nothing was written down at the moment the mismatch was visible on screen and could still be photographed cheaply.

Three costs usually arrive together. The first is time: a decoy page resolved by message can take days before dispatch, and the transit clock has not started. The second is return freight, which on a low-value accessory can approach the item price and is rarely worth paying. The third is the option value lost, since money held in an unresolved order is money not spent on a listing that could be verified.

Desk practice is to record four things for every row it follows: the page title as written, the item the seller photographs actually show, the option selected, and the date. Where a page will not settle into one of the four shapes, that ambiguity is written down as the finding. Unclassifiable pages are marked unclassified rather than judged, because a verdict built on a guess is worse than no verdict.

None of this requires a tool. It requires thirty seconds before paying, spent comparing three strings: the row description, the page title and the item in the gallery. Readers who do that report fewer surprises than readers who audit quality first and identity second, which is the order this desk recommends reversing. The protocol page sets out how the desk weights each of those three strings when they disagree. A written rule beats a remembered one, because the comparison has to be made while the page is open rather than afterwards.