What we found
The gap is wider than the conversation around dupes usually suggests, and it is not evenly distributed. A median saving of 83% sounds implausible until you look at what is driving it: the further up the price ladder the original sits, the larger the percentage becomes, almost mechanically.
26 of the 44 looks came in above 80%. Only 4 fell below 50% — and 3 of those 4 were trainers.
That is the most useful thing in the whole dataset, and it cuts against the headline. The category where the gap closes hardest is the one where construction, footbed and sole do the work, and none of those photograph. Where a saving is small, it is usually small for a reason worth respecting.
By category
| Category | Looks | Median look | Median cheapest | Median saving | Range |
|---|---|---|---|---|---|
| Clothing | 21 | £85 | £17 | 83% | 53–100% |
| Shoes | 13 | £120 | £27 | 78% | 37–98% |
| Bags | 9 | £495 | £45 | 89% | 63–99% |
| Accessories | 1 | £720 | £480 | 33% | 33–33% |
Accessories: One page. Reported for completeness; a median of one is not a median.
Bags show the widest gap and also carry the highest median original, which is the pattern in miniature. Shoes show the narrowest — and shoes are the category where our own writing most often says a cheaper version is a false economy, because construction and footbed decide comfort and neither of them photographs.
How we measured it
The inclusion rule was written down before the numbers were looked at, which matters — deciding what counts after seeing the result is how a study turns into marketing.
- Every published page in the library at 15 August 2026. Nothing was excluded for being unflattering.
- Alternatives are the ones each page actually shows a reader, with the original excluded.
- The cheapest is the lowest live price among those alternatives, in pounds.
- The saving is measured against the original's listed price on the date it was checked.
- Each page records its own price-check date; the window ran 12 June to 14 August 2026.
Every page behind these figures is public and individually checkable in the dupe library.
What this does not show
Six limits, including the one that makes our own headline number look worse.
One page reports a 100% saving, which is impossible
A saving of 100% would mean the alternative is free. It is not — this is a defect in how the figure is calculated against very high-priced originals, and it affects one of the 44 pages. The site already refuses to display any saving at or above 100%, so no reader has ever seen it, but the stored number is wrong and we are not going to quietly drop the row to tidy the range. The underlying calculation is being fixed.
The headline percentage depends on what we compared against
A saving is measured against the original's listed price, so the higher the anchor, the larger the percentage. Our biggest savings cluster on originals between £320 and £3,500, and the bags category — median original £495 — unsurprisingly shows the highest median saving of any category. Read the percentage as the distance between two prices, not as a measure of how close the two products are.
A listed price is not what people actually pay
Originals were priced at their listed price on the date checked. Anything bought in a sale, secondhand, or with a discount closes the gap, and a meaningful share of fashion is bought exactly that way. The savings here are against list, and list is the generous end.
A large saving is not evidence of equivalence
£3,500 to £35 is a price relationship, not a statement that the two objects are the same. The cheaper item is usually a different material, a different construction and a different lifespan. Our own writing on when a cheaper version is a false economy applies to our own library as much as anywhere.
The sample is small and it is curated
44 looks chosen by us, not a random sample of the UK market. They skew toward items people were already searching for, which is the whole reason they were made into pages. Treat this as a description of one library, not a survey of British retail.
Prices move
Every figure was captured between 12 June and 14 August 2026 and each page records its own check date. Stock sells out and prices change; some of the specific cheapest-alternative prices behind these medians will already be stale.
Questions
- How much cheaper is a dupe in the UK?
- Across 44 looks priced between June and August 2026, the median saving was 83% — a median original of £132 against a median cheapest alternative of £24. Half of all looks fell between 74% and 94%. The percentage is measured against the original's listed price, so it describes the distance between two prices rather than how alike the products are.
- Which category has the biggest savings?
- Bags, with a median saving of 89% — but that is largely because bags had by far the highest median original price at £495. Shoes showed the lowest median saving of the well-sampled categories at 78%, and also the single narrowest saving at 37%.
- How was this measured?
- Every published page in the DUPI library at 15 August 2026 — 44 looks and 191 alternatives from 109 retailers. Alternatives are the ones each page actually shows, with the original excluded. The cheapest is the lowest live price among them. Nothing was excluded for being unflattering.
- Does a big saving mean the products are the same?
- No, and this is the most important limit on the whole study. A saving measures the distance between two prices. The cheaper item is usually a different material with a different construction and a different lifespan, and one of our 44 pages carries a stored saving figure of 100%, which is a calculation defect rather than a finding.
Updates
- 15 August 2026 — First publication. 44 published pages, 191 alternatives, 109 retailers.
The figures above are a snapshot rather than a live count, so a finding stays quotable. When the library grows enough to change the answer, the study is rerun and a line is added here rather than the old numbers being edited away.
Using this
These figures are free to quote with attribution to DUPI and a link to this page. If you want a cut of the data we have not published, ask.