Kakobuy Spreadsheet Facts on One Page

The ledger compresses into a single line. It holds 195 entries extracted on 2026-09-29, drawn from 8 categories and 20 brands on one source marketplace, priced from $5.17 to $451.10 with a median of $31.76, carrying 2,806 images between them, of which 158 entries clear the five-image index threshold and 37 do not. Every other page on this site is an expansion of that line. This page states it once and attaches the three qualifications that keep it readable.

The first qualification is that this is one extraction, not a series. No second snapshot of the same fields exists, so no figure here describes a change over time, and none of the numbers should be read as a rate. The second is that the pool comes from a single source marketplace, so it describes what that marketplace listed on that date rather than a market. The third is that the status column counts fields, not quality: a row is verified when the record is complete enough to publish, which is a statement about the record and not about the item.

The page is grouped so a reader can stop at the group that answers the current question. Five figures cover scale. Three questions cover the three things readers ask most often. The last section names where the detail lives, including the vault at /vault/ and the question archive at /faq/.

Five figures carry most of the weight on this page, and each one decides something different. They come from the same extraction on the same day, so they can be read together without worrying about dates drifting apart.

Read the fifth row slowly, because it is the one that changes a published count. The 37 entries that stay out of the index are not defective items and were not judged. Their image groups hold fewer than five images, and the index layer admits a row only when its image group reaches five. The threshold is a publishing rule applied to a field, and it is stated in advance so the count can be reproduced by anyone reading the same pool with the same rule.

1. The pool count sets every share on this site. A share of 195 is a share of what has been recorded, not a share of the market.
2. The brand count limits brand comparisons. Twenty brands spread across eight categories means most brands appear in a handful of rows, and a brand average drawn from two rows is close to meaningless.
3. The median price is the middle value after sorting, not a typical order. An order holds several items and pays freight on top, so the median item price is a floor for the basket rather than an estimate of it.
4. The image total sets the scale of the QC record. A median entry carries 14 images, the largest group holds 46, and the smallest holds a single image.
5. The index split is the only place where a threshold changes a visible count: 158 entries in, 37 out, with the rule published rather than implied.

Three questions arrive more often than the rest. Each has a short answer with a number attached, and each answer is bounded by what the pool actually records.

How many entries does this spreadsheet hold? This ledger holds 195, taken from the eight category views of one source marketplace on 2026-09-29. Figures in the thousands appear in public discussion of agent spreadsheets, and those figures are not comparable with this one: they are usually quoted without a date, without a platform mix, and without a rule for removing duplicates. The number here is smaller because it is specific about all three.

What does an entry cost? The pool runs from $5.17 to $451.10 with a median of $31.76, and the category medians spread by a factor of roughly four, from $16.22 for Headwear to $68.68 for Shoes. A cost question is therefore a category question first and a price question second. Someone asking what an entry costs usually wants to know what a basket costs, and that requires the freight line as well, which is why the chargeable weight estimator at /instrument/chargeable-weight/ exists.

Is a record verified? Of the 195 entries, 158 are verified, none sit in the watching state, and 37 are unverified. The 37 are exactly the entries whose image groups fall below five images. Because the derivation uses two fields and one threshold, the count can be reproduced from the same pool without access to anything private.

The full table lives at /vault/, which lists all 195 rows with category, price, image count, source platform and status on one page. Eight category views sit beneath it, one per category, each carrying the count, the price range and the median for that group alone. A reader who wants to check a single figure on this page should start there, because the vault shows the rows that produce it.

Prose questions are collected at /faq/, which is where the recurring ones end up after they have been answered enough times to be worth writing down. Two instruments cover most cost follow-ups. The agent fee comparison at /instrument/agent-fees/ holds every rate as an editable input and excludes unverified fee fields from its totals rather than counting them as zero. The chargeable weight estimator at /instrument/chargeable-weight/ separates scale weight from volumetric weight and converts the difference into money, which is usually the largest single surprise in a first order.

Two of those six lines scale with the item subtotal and four do not, which is the shape most first orders get wrong. Commission and payment surcharge move with value. Domestic shipping moves with item count. Freight moves with weight and volume. Destination tax moves with the declared value against a threshold. A cheap item bought in quantity therefore pays more than its share of domestic shipping, and a light but bulky item pays more than its share of freight.

The habit worth taking from this page is to treat the median as a starting point and the range as the working figure. The pool maximum of $451.10 sits in Accessories, a category whose median is $26.77, so the same category contains both the cheapest entry in the pool at $5.17 and the most expensive one. A summary that reports only the median would hide the entire spread that makes the category interesting.