Legit covers three separate claims and usually only one gets tested
Three different claims hide behind one word. The sheet is current. The sheet is honest about where its values came from. The sheet is not a route for someone else to take money from you. A sheet can pass the first two and fail the third, and no amount of accuracy in a status column fixes that.
Split the question before answering it. Everything in this check is about reading a document, not about accusing anyone of anything. You are testing whether a source can carry a decision, and the answer is often that it can carry a plan but not an authorization.
- Freshness: can you tell when each row was last true?
- Provenance: can a value be traced back to a platform record?
- Correction path: if a value is wrong, is there a route to fix it?
- Payment boundary: does the sheet ask for money, or only show records?
- Scope: does it claim to settle things no sheet can settle?
The five-point check below takes about ten minutes on a 200-row sheet. Run it before you copy anything, and keep the notes; the same five points work on the next sheet you look at.
Point one: date the sheet before you read a single status word
Find the newest and the oldest copied-on date and write the spread down. A three-day spread means somebody is active. A six-week spread means you are reading an archive. No copied-on column at all is itself a result: undated rows cannot be ranked against each other, so the sheet cannot tell you which item moved today.
Freshness is not a matter of trust. It is arithmetic on two dates, and it is the only one of the five points you can settle in under a minute without knowing the maintainer.
Points one and two take about four minutes together. If the sheet fails either, stop there and treat the remaining reading as orientation rather than evidence.
Points two and three: trace a value, then try to break it
Pick three rows at random, not the three at the top. For each one, count whether it carries a platform item ID or an order reference, or whether it is just a product name and a status word. Rows with references can be re-checked; rows without them cannot. Write down how many of the three you could trace.
Then look for a correction path. Sheets that expect to be wrong usually say so: a revision note, a comments column, an owner name. A sheet with 200 rows, no owner, and no correction route is not lying to you, but every error in it is permanent, and errors accumulate in the direction of optimism.
| Rows traced | What it means | How to use the sheet |
|---|---|---|
| 3 of 3 | Values are pasted from records | Usable as an index, still re-check before acting |
| 2 of 3 | Mostly copied, some typed notes mixed in | Trust referenced rows only |
| 1 of 3 | Mostly free text | Reading for orientation, nothing more |
| 0 of 3 | No traceability at all | Treat status words as commentary |
Points four and five: the payment boundary and the claims that overreach
A sheet that shows records and a sheet that collects money are two different objects, even when they look identical. If a row contains a payment link, a personal account, or a request to send funds outside the platform you already use, stop and treat the whole document as an offer rather than a reference.
Overreach is the quieter failure. Any row that promises a duty amount, a tax outcome, or a customs result is claiming authority that belongs to a customs authority, not to a maintainer. Import charges are set by the destination country and assessed case by case, so a fixed number in a cell is at best an old average and at worst a guess.
No sheet is quoted here as an example, and no maintainer is named. Where a value cannot be checked against a record on the day you read it, this page calls it unchecked rather than reliable or unreliable.
- Payment requests inside a sheet: treat as a stop signal.
- Fixed duty or tax figures: check against the destination customs authority.
- Claims about service windows or fees: ask for the platform page, then read that page.
- Anything phrased as certainty about an outcome you do not control: lower its weight.
The check people skip: does the sheet actually cover your route
A sheet can pass all five points and still be the wrong document for you. Coverage comes in three parts: which agents appear in it, which source marketplaces those agents were listed for, and which destination countries the rows describe. A sheet built around one agent and two marketplaces tells you nothing about a third marketplace you used last week.
Count the agents listed and compare that number with the agents you actually use. Then look for the destination side. A status column copied from another buyer account will not describe your customs route, your carrier, or your delivery address, because none of those belong to the person who wrote the row.
- List the agents you have used in the last 90 days, then check each one appears.
- Check whether the source marketplaces in the sheet match your own orders.
- Check whether rows mention a destination country at all.
- If your route is absent, the sheet cannot speak to it, however accurate it is elsewhere.
This is the point where copying a general sheet usually goes wrong. People copy a good sheet, then act on it for a route the sheet never described, and the failure looks like bad advice when it was really a scope mismatch.
When the sheet you found is a copy of a copy
Popular sheets get duplicated. If three maintainers each keep a version of the same list, you may be reading the third copy, and staleness compounds along the chain: a 30-day refresh copied into a second file and then a third can be two months old while the file itself was saved yesterday.
You can often spot it by fingerprint. Identical column names, identical typos, identical row order, and a saving timestamp far newer than the oldest copied-on date inside the file. That combination means the file was re-saved, not refreshed.
- Find the earliest version you can reach and check whether it is still maintained.
- Compare the oldest copied-on date with the file save date.
- Ask the sharer which file they copied from, and when.
- Prefer one maintained file over four copies, even if the copy looks tidier.
A short note of your own beats all of this. One line per item, with the date, the record reference, and the stage it was in, is a document you can defend, because you know exactly who typed it and when.
Grading your five answers, and what each grade lets you do
Add up the passes. One or two passes out of five means the document is advertising with a table layout. Three or four means it is a usable reading source. Five passes means it is a decent index, which is still not the same as authority, because the record you act on is the platform record.
| Grade | Passes | What you may do with it |
|---|---|---|
| Read for orientation | 1 to 2 | Skim it, verify everything before acting |
| Use as a reading source | 3 to 4 | Plan your week from referenced rows |
| Use as an index | 5 | Track items, still re-check eligibility live |
Whatever the grade, two things stay outside the sheet: whether a specific action is still open on your order, and what your parcel will cost at the border. The first lives in your order record. The second lives with the customs authority of the destination country, as outlined in the [platform checks topic](/table/legitimacy/).
If the sheet came recommended in a thread, take the thread claim and run these five points on the sheet itself. Unanswered cases like this are collected under [open doubts](/doubts/), and the [agent selection stage](/steps/pick-agent/) is where the result actually changes what you do next.
Worked example: sample 3 rows from any maintained sheet and re-check the same items against live records; with a 3-row sample a result of 2 out of 3 still matching means the sheet is roughly within one refresh cycle, while 0 out of 3 means you should stop reading its status column for eligibility. (stated basis: Worked example)
This note explains the mechanics; the list is where the items are.