Index rebuilt on the new catalogue snapshot
307 rows re-pulled from the partner catalogue, 41 dead listings removed, 12 QC sets appended to the log.
2027edition
Annual index · rebuilt 26 October 2026
Kakobuy spreadsheet 2027
The year edition is not a reprint of the homepage. It is the status report: which rows survived the re-verification, which batches changed, and what we stopped trusting. Below is the changelog, the shelf table, and the four questions that decide whether a row stays.
Changelog
Shoes split by batch family
Shoe rows now carry the factory batch tag where the seller states it, because two listings with the same photo can ship from different batches.
Shipping estimator rewritten
Volumetric divisor now depends on the line (6000 economy/standard, 5000 express) instead of a single fixed number.
QC finder shipped
Search by SPU number, keyword or category. The tool reads the same index the pages do, so results cannot drift.
Coupon ledger started
Codes are now dated and marked reported / announced / expired instead of being reprinted forever.
Eight shelves, measured
Row counts are the number of live entries in this snapshot, not the number of products the seller has. Median price is the middle row of the shelf, in USD as listed.
Shelf tablesnapshot 2026-10-26
| Shelf | Rows | Median | QC sets | What we check first |
|---|---|---|---|---|
| Shoes | 48 | $47.81 | 8 | outsole alignment, midsole glue, tongue label, box label |
| Hoodies & sweaters | 7 | $25.43 | 3 | fleece density, embroidery back, cuff stretch, drawcord metal |
| T-shirts | 48 | $25.36 | 8 | print edge, collar ribbing, side seams, shoulder taping |
| Jackets | 48 | $50.72 | 8 | zip pull, baffle stitch, lining seams, hood structure |
| Pants & shorts | 12 | $50.86 | 3 | inseam, rise, hardware, hem finish |
| Accessories | 48 | $84.76 | 8 | stitch per cm, edge paint, hardware weight, lining |
| Headwear | 48 | $16.87 | 8 | brim shape, embroidery, strap hardware, sweatband |
| Other stuff | 48 | $37.14 | 8 | box condition, plug type, battery restrictions |
How a row survives into 2027
A row survives if it resolves on two separate weekly snapshots, its price has not moved by more than 30%, and any QC set on file agrees with the title and size. Fail twice and the row is retired — the changelog above records the removal instead of hiding it.
1
It resolves twice
Opened on two separate snapshot dates. A listing that 404s once gets flagged; twice, it is retired.
2
The price holds
A move of more than 30% since the last snapshot means the seller changed batch or bait-switched. We re-check before keeping the row.
3
The evidence agrees
If a QC set exists, the size and colour in it must match the row. Mismatch means the row is lying about something.
What "verified" means here: the row resolves to a live product page, the variants are purchasable, and the price matches the snapshot. It does not mean we inspected the physical item — only the warehouse can do that.
Six rows we kept, and why
![Stussy Hoodie [15 styles]](https://kakox.com/api/static/products/WEIDIAN_7778396676_1_284x284.webp)
![Dior Heels [2 styles]](https://kakox.com/api/static/products/WEIDIAN_7775361273_1_284x284.webp)
Questions about the 2027 edition
What is different in the 2027 index compared with last year?
Three things: every row now carries a resolution status, the shipping estimator uses per-line volumetric divisors instead of one fixed number, and the QC log is organised by check type rather than by date. The catalogue itself is the same partner catalogue, re-pulled weekly.
Why publish an annual page at all?
Because "kakobuy spreadsheet 2027" and "kakobuy spreadsheet" are not the same query. The head term wants an entry point; the year term wants a status report — what changed, what is current, what is stale. One page cannot serve both without being vague.
How do you decide which rows survive into the new year?
A row survives if it resolves, if the variants are still purchasable and if the price has not moved by more than 30% since the last snapshot. Rows that fail two snapshots in a row are retired and noted in the changelog.
Is the row count a quality signal?
No, and we would rather say so. A 30,000-row list that nobody checks is worse than a 300-row list where every entry resolves. Count how many rows in a spreadsheet still work before you judge it by its headline number.