How to read a finds list in five steps
Five steps from list to order
The method below assumes a list you did not write, which is the harder case: you have to audit someone else's fields rather than your own.
Five steps, and only the second one costs real time. The rest are habits that take seconds once they are automatic.
- Scan for rows that carry fields, not rows that carry pictures
Fifteen seconds per screen. If a row has no source platform and no measurement, it cannot be verified later and is not worth shortlisting regardless of how good the thumbnail looks.
- Open the source for the two or three rows that survived
Check the price and the variant still exist. This is the step that catches stale lists, and it is cheap because you are only opening a handful of pages.
- Read the measurements against something you own
A chest width in centimetres compared with a garment in your wardrobe beats any size label. Where a listing gives a range, treat the middle of the range as the likely truth.
- Estimate weight before you commit
Weight decides the shipping band, and shipping is often the largest single line. A rough figure from a similar item is enough to catch the case where shipping exceeds the item.
- Order, then hold the QC window open
Photographs arrive at the warehouse. That is the moment to compare against the source image you recorded in step one — which is only possible if you recorded it.
What to skip at each step
Skipping well is most of the skill. Each row below names something that feels productive and is not.
| Step | Skip this | Because |
|---|---|---|
| Scanning | Comparing thumbnails across rows | A thumbnail is the least verifiable field in the row and the one the compiler chose |
| Verifying | Opening every row | Most rows will not survive the first filter; opening all of them spends the budget before the decision |
| Measuring | Trusting the size label | Labels differ between factories; the numeric measurement is the only portable fact |
| Weighing | Estimating from the price | A cheap bulky item and an expensive small one invert the intuition entirely |
| Ordering | Requesting extra photos before comparing | Compare against the source first, then request only the angles that are actually missing |
Where this method breaks down
Three cases where the five steps produce a confident answer that is wrong.
- When the listing is a variant set
A single row covering forty colourways has one price and forty answers. The steps still work, but they apply to the variant you pick rather than to the row you saved.
- When the seller changes the listing
Mixed catalogues get renamed and repriced. A verified price is verified on a date; if the row is a month old, verify again rather than reasoning about it.
- When measurements are given as a range
A range often means the seller is covering several production runs. Plan for the middle, and expect the possibility of either end.
Knowing where a method fails is what makes it usable. The three cases above are why the verification step exists at all.
Reading questions
Four answers about the method.
How long should the five steps take?
What if the list has no source platform column?
Should I verify every price?
Do I need the printable checklist?
Step two needs a source to open
The index records the source platform on every row, so verification is one click rather than a search. 11,414 listings, each with its platform recorded.