Three numbers decide what you pay to move a parcel: the item price, the service fee, and the freight. Two of them are printed on the product page. The third one — the one that quietly gets bigger after the parcel reaches the warehouse — is built on a figure the product page never shows: the weight. On a single 1688 listing we pulled apart, the seller had registered twelve per-SKU weights between 0.911 and 0.985 kg. A quoting tool that has no weight to work with does not stop; it substitutes a default, usually half a kilogram, and the estimate comes out 45% light. This post is about where that number really lives, which source to trust when two of them disagree, and the arithmetic that decides whether you are billed on weight or on volume.
1. Every freight quote rests on one number, and that number is usually a default
Freight is priced by the kilogram, so the weight is not a detail of the estimate — it is the estimate. Which is why the most consequential field in the whole calculation is also the one no product page displays. When a comparison tool cannot read a weight, it has three options and only ever uses one: substitute a default. Half a kilogram is the usual choice, because it is a reasonable average for small consumer goods and because nobody complains about an estimate that comes in low. On the listing we measured, that default was not a rounding error. Twelve SKUs, twelve registered weights, spread from 0.911 to 0.985 kg. The lowest of them is 82% heavier than the assumed 0.5 kg; the heaviest is nearly double. And because agents bill on weight bands, a 45% light estimate does not stay 45% light at checkout — it can cross a band and add a step the comparison never warned you about. The ranking is the casualty. When two agents sit within a few yuan of each other, the weight assumption is a bigger input than the price difference, which means the cheapest agent on screen is not reliably the cheapest agent in your account.
2. Three places a 1688 weight can come from — and which one to believe
The first source is the seller dashboard, and it is the only one worth building on. When a merchant sets up a listing, the backend asks for a per-SKU piece weight and pack dimensions — grams plus length, width and height in centimetres. That record is structured, it is required for shipping calculations, and 1688 delivers it to the page inside the page data rather than in the visible layout, which is exactly why tools that scrape what a human can see walk straight past it. If you want to spot it yourself, the field is called pieceWeightScaleInfo and it carries one entry per SKU. The second source is the free-text detail a seller occasionally types into the description — a line reading “单件重 0.9kg”. It is human-readable, it is optional, and on most listings it simply is not there. The third is the specification table, where weight sometimes appears among a dozen unrelated attributes, in grams on one listing and kilograms on the next. When two of these disagree, believe the dashboard record: it is the one the seller had to fill in for the platform own shipping arithmetic, and the only one that is per-SKU rather than per-listing. The order matters. Read the structured record first, fall back to the visible text, and only then admit you do not know — the failure mode to avoid is filling the gap with a confident default nobody can see.
3. Twelve SKUs, twelve weights: a colour swap moves your freight
A 1688 listing is not one product, it is a matrix. Colour, size, and bundle quantity each generate their own SKU, and each SKU carries its own registered weight, because the seller fills that form row by row. On the listing we measured, the twelve rows spanned 0.911 to 0.985 kg — an 8% spread inside a single product page. Eight percent does not sound like much until you remember what it multiplies: a long-haul freight rate applied to a parcel whose landed cost sits within a few yuan of its nearest competitor. Quote the lightest SKU and take an order for the heaviest, and you have understated the shipment by design — the shortfall lands on your margin and on nobody else. The reverse error is just as expensive: quoting the heaviest variant for a customer who buys the lightest makes your offer look worse than it is, and on a comparison page that costs you an order you should have won. So the rule we adopted for our own calculator is deliberately asymmetric. When the weight varies across SKUs, take the maximum. An estimate that comes in slightly high is survivable, because the surplus visibly returns to the buyer. An estimate that comes in low is not, because the difference leaves the seller pocket silently, and the first thing that follows a surprise charge is a support message.
4. Verifying it yourself in three steps — and when volume beats weight
You can verify a weight without any tooling, and there are exactly three places to look, in one fixed order. Open the specification block first and scan for a per-item weight, because that is the seller own entry. If it is silent, open the SKU selector and step through the variants — a few listings surface the weight only once a variant is picked. If both come up empty, ask in 旺旺 and keep the reply, the same way you would ask which warehouse an order actually ships from. Then run the arithmetic that catches almost everyone: compare the real weight against the volumetric one. Multiply length by width by height in centimetres, divide by the divisor your carrier line uses — 6000 on most express lanes, 8000 on some economy ones — and take whichever of the two figures is larger. Our sample parcel measures 40 × 32 × 4.5 cm, which is 5,760 cubic centimetres. Divided by 6000 that is 0.96 kg, divided by 8000 it is 0.72 kg, and since the registered weight is 0.985 kg, this parcel is billed on what it weighs, and only barely. Enlarge the same box to 50 × 40 × 30 cm and you are at 60,000 cubic centimetres, which at a divisor of 8000 is 7.5 kg against a real weight of perhaps one kilogram. Now you are paying for the air. That flip is why the volumetric divisor is not an accounting footnote but the single setting that decides whether a light, bulky product is a bargain or a trap — and it is worth checking before you commit to a category, not after the first parcel has been billed.
FAQ
Why does the same product cost more to ship in a different colour?
Because each colour is its own SKU and carries its own registered weight. Sellers fill the piece-weight form row by row, and the rows are under no obligation to agree: on the listing we measured, the twelve variants ran from 0.911 to 0.985 kg. Different fabric, a longer zip, a bundle that adds one accessory — small physical differences become a different billing weight, and once a parcel crosses a weight band the freight steps up with it. That is also why a quote built on the lightest variant can be wrong by the time the customer actually picks a colour.
The seller never filled in a weight. What now?
Fall back in order and stop when you run out of sources. First the specification table, then the visible description text, then 旺旺. If all three are silent, you do not have a weight — and the honest output is an estimate that is labelled as an estimate, not a number dressed up as a measurement. What you should not accept is a tool quietly inserting its own default and showing you a total that looks exactly like every other total on the page. That is how a half-kilogram assumption ends up deciding which agent you buy from.
Does the 1688CMP extension verify the weight for me?
It reads what the listing already carries, in the order above: the seller per-SKU piece weight and pack dimensions from the page data first, the visible weight text second. Both are the seller own numbers, not ours. What the extension will not do is invent a weight when the seller left the field empty — in that case you get a default and should read the total as an estimate. The line worth keeping in mind: it prices the international leg on the SKU you actually selected, and the domestic leg into the agent warehouse remains yours to check.