Ordering
Our testing team places an order for every featured product at the listed price, using an ordinary buyer account from Rotterdam. We record the order date, the quoted lead time and the price at checkout, then log the actual delivery date when the parcel arrives.
Arrival and bench checks
Parcels are opened on camera. We weigh and measure the item, check stitching and hardware where relevant, photograph labels, and note packaging quality and any customs paperwork. Where a product is offered in sizes, we order two sizes and record the fit difference rather than averaging it.
Agent timing
For sourcing-guide work we send the same two questions to three agents and time the first substantive reply, then check whether the answer matches the catalog entry. Response times are reported as a range across three attempts, not a single best case.
Comparison tables
Each review sits in a table with at least two alternatives checked on the same day. Columns cover price, delivery time, link stability, packaging condition and agent responsiveness, so a score can be read against its neighbours.
Scoring
Scores use our five-point scale, with hands-on product scores published between 4.0 and 5.0. We explain each score in prose: 5 excellent, 4 good, 3 adequate, 2 poor, 1 very poor. A score is never an average of columns alone; the reviews editor writes the reasoning and the editor in chief checks anything above 4.5.
Retests
Featured listings return to the bench every nine months, and links are re-checked quarterly. If a source link redirects or an agent changes its reply pattern, the affected page goes back into the test queue before the score stands.