Topic · 7 entries
Platform checks
How to tell a maintained sheet from an abandoned one, and what "legit" can and cannot mean.
Every entry, answer first
Open an entry to see the basis it rests on, the moves it implies, and the way people usually misread it. Every entry names the stage it bites at.
How do I tell a spreadsheet that is still maintained from one that was abandoned?Look at the newest dated row rather than the banner: a sheet is as current as its most recent entry.
Worked example Subtract the date of the newest row from today and read the result as the age of the sheet. A list whose newest entry is 3 months old, under a header claiming daily updates, tells you which of the two to trust. Then spot-check 3 links: a maintained sheet usually has working links near the top, while an abandoned one has a cluster of dead or redirected ones. The subtraction takes a minute and it is the only maintenance test that uses evidence rather than tone.
What to do
- Find the newest row with a date and note how old it is.
- Open three links from different parts of the sheet and check that they resolve.
- Treat a wide gap between the header claim and the newest row as your answer.
The misread is judging a sheet by how professional the header and the layout look.
Bites at: stage 1 — Pick an agent
Does legit mean the items on a sheet are authentic?No: the platform states that selling non-genuine items is prohibited, and buying them can cost you the order or the funds.
Public rule The published policy states that selling non-genuine items is strictly prohibited, that a breach can lead to loss of purchasing rights, withheld funds and destroyed stock, and that where a purchase of non-genuine items is detected the platform may refuse the order, suspend the account and withhold the money (per Kakobuy help centre ?id=125, captured 2026-10). That is a rule about the platform’s own exposure rather than a guarantee about a listing, and the chain from link to stage 5 photographs never tests what an item is made of. Check the current rule before you read a well-kept list as reassurance about the goods.
What to do
- Read legit as maintained and reachable, and nothing more.
- Check the newest row date before you use any link from the sheet.
- Treat the risk on the goods themselves as yours to decide, given the published policy.
The misread is reading a well-kept list as a quality guarantee about the goods behind the links.
Bites at: stage 2 — Hand the link over
If a spreadsheet lists prices, can I trust the numbers in it?Treat any listed price as a snapshot of the day it was collected; the listing at checkout is the figure that binds.
Worked example Every price in a third-party list has a collection date attached to it, and the difference between that date and today is the age of the figure. A price collected on the 1st and read on the 20th describes a listing that has had 19 days to change, and the published total only freezes the item price when the agent pays at stage 3. Compare the listed figure with the one confirmed at checkout, and treat a difference as normal rather than as evidence of anything.
What to do
- Find the collection date next to the price before you use the figure.
- Check the live listing for the same item yourself.
- Write down the price the agent confirms, and compare the two afterwards.
The misread is reading a stale price as a promise, and a changed price as a sign of a scam.
Bites at: stage 3 — Pay for the item
How do I tell whether a review of an agent was paid for?You usually cannot from the text; you can only sort reviews by how many checkable details they contain.
Worked example Count the specific claims in a review instead of weighing its tone. A review that names a weight, an order status and a date gives you 3 things to test against your own experience; one that lists only adjectives gives you nothing to check, whether it was paid for or written in good faith. Across 10 reviews, the ratio of checkable to uncheckable ones is a rough guide, and it is the only guide available without knowing who wrote them.
What to do
- Count the numbers and dates in a review before you weigh its opinion.
- Look for the same complaint repeated by different authors, which is harder to stage.
- Test one specific claim yourself on a small order before trusting the rest.
The misread is assuming a negative review is honest and a positive one is bought.
Bites at: stage 1 — Pick an agent
What should I check before paying an agent I have never used?Check that the account page, the payment page and the support channel share one domain, and that a person answers.
Flow dependency The payment leaves at stage 3, and no later stage reverses it: from the warehouse stage onward the process is about the item, not about the money you sent, and the published policy even allows the platform to refuse an order or withhold funds where non-genuine goods are involved (per Kakobuy help centre ?id=125, captured 2026-10). One domain across the account page, the payment page and the support address is the basic test, and a mismatch between two company names on the same page is a stop signal. Send one specific question to support first and see whether a person replies with a specific sentence.
What to do
- Compare the domain on the account page, the payment page and the support address.
- Send one specific question to support and read the reply for specifics.
- Start with a single low-value item rather than a full order.
The misread is judging an agent by its landing page and by how fast it answers a first message.
Bites at: stage 3 — Pay for the item
Can I split one order across two agents to compare them?You can, but you lose the merge: two warehouses mean two parcels and the fixed charges paid twice.
Flow dependency Each agent holds its own warehouse, and merging at stage 6 only combines items that are in the same building, since each agent’s storage window and return window are counted inside its own system. Splitting one order across two agents therefore produces at least two international parcels, each with its own fixed charges and its own shipping request, and the published per-parcel declaration fee is one of them. That is a reasonable price for a comparison test on one cheap item; it is an expensive way to order a set of items that would have merged into one box.
What to do
- Decide whether the test is worth two parcels before you split the order.
- Keep the item you are testing small and cheap.
- Compare the two bills line by line at the end, not the two totals.
The misread is expecting to compare agents on price while changing two variables at once.
Bites at: stage 1 — Pick an agent
What if the links in a sheet stop working?Link rot is normal on marketplace listings, so test a few links before you trust the rest of the sheet.
Flow dependency A sheet stores links, and a listing link depends on a marketplace URL that can change when a seller relists an item. The order of events matters: you choose a link at stage 2, the agent opens it at stage 3, and by then a dead link is a dead end for that row rather than for the whole sheet. Test three links from different sections: if one is dead and two resolve, the sheet is usable; if all three are dead, you are reading an archive.
What to do
- Open three links from three different sections before you rely on the sheet.
- Search the marketplace for the same item yourself when a link is dead.
- Report the dead row to whoever maintains the sheet, with the date.
The misread is concluding the whole sheet is fake because one link in it no longer opens.
Bites at: stage 2 — Hand the link over