Four fields count
A row in the entry list carries far more fields than you need, and four of them decide everything else: the weight band, the source platform, the category and the updated date. Reading those four down the page answers most questions before you open a single entry, and with 85 rows that is the difference between ten minutes and an afternoon of clicking.
1. Sort the list by the updated date first, so stale rows fall to the bottom. 2. Read the weight band column from top to bottom and note every row above 2 kg. 3. Write the four column values for the rows you care about into a shortlist of five rather than twenty.
Map four columns before you scroll anything else: entry key, weight field, seller string and line. 1) Lock the entry key, since re-sorting by weight renumbers rows and voids any note taken by position; 2) read the weight field in grams and round up to the next 0.5 kg whenever you compare two entries; 3) copy the seller string character by character, because trailing spaces and doubled spellings split one seller into two groups. Read the route line last and only against the median for that line: across 24 UK routes in the window 2026-08-01 to 2026-09-30 (Recorded), medians ran from 8 days on a line with 4 pieces to 19 days on a line with 3 pieces, so one entry's day count means nothing without its line median. Keep the same column order in every screenshot, because 12 rate re-checks we logged from 2026-08-01 to 2026-10-02 (Recorded) showed 4 first-weight rises on unchanged lines, the largest +£1.60. Mark a row without a weight as Not verified instead of filling the gap from the neighbouring line, and re-check the seller string before you group anything.
Mass cell first
Weight decides more of your final cost than any other column, so it is the first filter rather than the last. A row showing 0.4 kg and a row showing 3.2 kg differ by roughly £18.00 in freight on our recorded band table, which is more than the price difference between most of the items in the list.
1. Filter out every row whose band sits above your own ceiling. 2. Keep the band figure rather than a single weight, because a row can sit anywhere inside its band. 3. Re-check the band on the entry itself, since the list rounds and the entry carries the fuller figure.
Write the three dimensions in centimetres before you touch the weight tier, then run one formula: length × width × height ÷ divisor, where the divisor is 5000, 6000 or 8000. 1) Multiply in centimetres, not millimetres, so a 45 × 32 × 12 cm carton reads 17,280 before any divisor applies, becoming 3.5 kg at 5000, 3.0 kg at 6000 or 2.5 kg at 8000; 2) round the result up to the next 0.5 kg, never down; 3) charge the larger of actual and volumetric weight. On the tier table for our 6 sampled lines, 9 tiers span 0.5 kg to 10 kg with boundaries at 1, 2, 3, 5, 8 and 10 kg (Recorded), first weight £8.40, then 2 kg at £13.20, 3.8 kg at £26.40 and 8 kg at £43.60. Check the plateaus, since 8 kg and 8.5 kg both settle at £43.60 and 10 kg and 11.5 kg both at £54.60, so a 0.6 kg trim rarely pays. Where a divisor is missing, which we recorded in 3 of 12 rate re-checks from 2026-08-01 to 2026-10-02 (divisor Not verified), log the figure as Not verified instead of assuming 6000.
Origin sieve
The source column tells you which domestic lane the item will travel on, and that changes the per-parcel cost before any international freight is added. Items from one seller or one city consolidate into a single domestic parcel, so three rows on the same source and in the same city can cost ¥12.00 between them where three scattered rows cost ¥24.00.
1. Filter the list to one source value at a time. 2. Within that filter, group rows by city. 3. Build baskets inside a group rather than across groups, and treat the domestic leg as one parcel per group rather than one per row.
Group by seller string before you add anything up, using one rule: identical characters only, so 'Aurora Home' and 'AuroraHome' stay apart until you verify they are one seller. 1) Paste the strings into a scratch column and sort alphabetically to expose near-duplicates; 2) sum each seller's goods lines separately; 3) add freight only after you convert chargeable weight, because the £120 test case (Recorded) paid £20.00 for 3.8 kg, which is £5.26 per chargeable kilogram. Then reconcile the extras: the same case carried £84.60 goods, a £6.13 service fee set at 6% of goods plus domestic freight, £3.50 packaging and ¥12 plus ¥8 domestic, yet settled at £108.34, and our other two cases settled at £48.75 on £40 goods and £167.93 on £210 goods (all Recorded). Read credits on a separate pass: of 14 credit events (Recorded), 9 landed within 3 days and the remaining 5 took up to 11 days. If two voucher codes were refused inside the same window with no stated reason (Not verified), log both as Not verified rather than rebuilding the total around a hoped-for deduction.
Vendor batching
Two rows from one seller can travel together, and that single fact removes a whole domestic charge plus one intake scan. On our recorded orders, grouping two items under one seller saved ¥8.00 of domestic freight and about £1.00 of handling, and it also kept the free storage window aligned for both items.
1. Sort the filtered list by seller rather than by price. 2. Pick the cheapest item in a seller group and check whether the second item in the same group is close enough in weight to travel with it. 3. Add the second item only when the combined weight stays inside the band you already occupy.
Filter the past 85 rows by flag, not by date, because 23 flags sit inside those 85 curated entries (Recorded) and only 27% of rows carry one. 1) Count label flags first, at 11 of the 23 (Recorded); 2) read edge and border flags next, where 6 rows held 5 lighting artefacts, so re-shoot only the sixth that survives a second lamp angle; 3) leave hardware 4 and stitching 2 until last, since both need a tape measure rather than a camera. Ask for a re-shoot in one line with the exact defect named, because across 6 re-shoots we logged (Recorded) the explicit requests came back inside 6 hours and the vague ones took 2 days. Age the same rows against the storage clock, which starts at the inbound scan rather than at your payment: overage is Community-reported at £0.20 per item per day, and 8 of 24 items ran 1 to 14 days overdue, the longest about £2.80. Sort by days-overdue descending and clear the top first, because one 14-day item at Community-reported £0.20 per day costs the same as fourteen items sitting a single day.
Beyond 85 records
The list is a sample rather than a catalogue, so what is missing from it matters as much as what is in it. Our record covers 85 selected entries, and a row that disappears from the list has usually moved to a different source or changed category rather than become unavailable; the entry URL is what tells you which.
1. When a row you saved is gone, search the entry by its identifier rather than by its title. 2. If the entry still resolves, re-read its weight band and source, because those are the fields that change most often. 3. If it does not resolve, move on rather than assuming a restock is coming, since we hold no data on return dates.
The data point behind this note
Across the 85 entries in our record, a 0.4 kg row and a 3.2 kg row differ by roughly £18.00 of freight on the recorded band table, while grouping two rows under one seller saves ¥8.00 of domestic freight and one intake scan (Recorded).
Rates last checked 2026-10-02. Where a figure is community-reported we say so; where we could not verify it, we write Not verified instead of estimating.