Sourcing is not one activity. It is four quite different businesses that happen to share an output, and each has a distinct cost structure, time cost and failure mode. Picking the wrong one for your capital and schedule is a more expensive mistake than picking the wrong brand.
The classic entry point: low capital, no minimum order, and you inspect every item before paying. The costs are time and variance. You might spend three hours and find two items worth listing, and you cannot plan around that — the stock is whatever was donated that week.
It scales badly on purpose. The moment you need forty items a week, the hours required grow linearly and the good shops get picked over by everyone else doing the same thing. It remains an excellent way to learn what sells, because the feedback loop between judgement and outcome is short and cheap.
Buying by the kilo or by the pallet inverts the trade-off: cheap per item, near-zero selection time, and you cannot see what you are buying. The economics work when your per-item cost is low enough that a large share of unsellable stock still leaves you ahead — and they collapse quietly when a supplier's grading slips.
The real risks here are commercial rather than operational. Supplier quality varies between lots from the same seller, sorting a bale is genuinely physical work, and you need somewhere to put several hundred garments. Start with the smallest lot a supplier will sell you, regardless of what the per-kilo price does at volume.
Buying new stock at markdown is the most predictable channel: known condition, known sizes, full size runs, and you can reorder what works. The margin is thinner because everyone can see the same clearance price, and the window is short.
This channel rewards knowing the resale value of specific models cold, because the decision has to be made in minutes and there is no negotiating. It is also the channel where a per-model buy-below price earns its keep most obviously — the question is never whether the brand is good, it is whether this particular price beats your number.
Buying underpriced listings on one platform and reselling them on another, or in another market, is the lowest-friction channel and the most competitive. No travel, no storage before purchase, immediate scale. The catch is that everyone else can also see the listing, and the good ones are gone in minutes.
It works when you have an information edge — you know a model's real value and the seller does not — and it fails when you are simply refreshing the same feed as three hundred other people. Note that the edge has to come from knowing the item, not from geography: we measured the five EU Vinted domains and found most listings appear on several of them at an identical price, so buying in one country to sell in another mostly does not work. Chapter 12 has the numbers.
Charity and secondhand, almost always. The capital at risk per mistake is a few euros, you see every item before paying, and the fast feedback teaches you what sells far quicker than reading about it. Move to bales or clearance once you can predict sale prices without looking them up.
They can be, once you have both the storage and the pattern recognition to sort quickly. They are a bad first move because you cannot yet tell a good lot from a bad one, and the loss on a bad first pallet is large enough to end the experiment.
Resale IQ turns a Vinted listing into one answer: BUY, WATCH, or SKIP — with buy-below price and best sizes.
Try Resale IQ →