Research record · 18 September 2026
What we refused to publish, and why
Product pages on this site have blank cells where a specification should be, and some retailers are missing from our price lists entirely. This page is the reason. Everything below comes from the same research file that fills the rest of the site, so nothing here is a claim you have to take on trust — each row names the sources that disagree, and you can open them yourself.
- Figures withheld
- 12 across 6 models
- Retailer links withheld
- 24 across 10 retailers
- Widest price gap
- 84% Roborock Saros 10
- Makers contradicting themselves
- 3 of 12 disputes
Finding one
The same robot, 84% dearer depending on the country
The Roborock Saros 10 is sold by Roborock itself in every market below. The cheapest and the dearest national store differ by £585 on the same machine, from the same manufacturer, on the same day. It is not a story about one country being cheated: which market wins changes model by model, and Britain is the cheapest for two of the 7 robots we can price in more than one country.
- United Kingdomcheapest£699.97
- Germany899,00 €≈ £771
- France949,00 €≈ £814
- Spain1.499,00 €≈ £1,285
Every model we can price in more than one country
Manufacturer stores only. Euro prices are converted to pounds at the European Central Bank euro reference rate of 16 September 2026 for comparison, and shown in their own currency first.
| Model | Cheapest market | Dearest market | Gap |
|---|---|---|---|
| Roborock Saros 10 | United Kingdom · £699.97 | Spain · 1.499,00 € ≈ £1,285 | 84% £585 |
| Dreame X50 Ultra Complete | France · 729,00 € ≈ £625 | United Kingdom · £999.00 | 60% £374 |
| Roborock Qrevo Curv 2 Pro | Germany · 699,00 € ≈ £599 | Italy · 1.099,00 € ≈ £942 | 57% £343 |
| Roborock Qrevo S Pro | France · 389,99 € ≈ £334 | Italy · 599,99 € ≈ £514 | 54% £180 |
| Roborock Saros 20 | United Kingdom · £998.99 | Italy · 1.499,00 € ≈ £1,285 | 29% £286 |
| eufy Omni C28 | Germany · 599,00 € ≈ £514 | United Kingdom · £649.00 | 26% £135 |
| Dreame X60 Pro Ultra Complete | Germany · 1.499,00 € ≈ £1,285 | United Kingdom · £1,299.00 | 1% £14 |
Rates move and promotions end, so a gap here is a snapshot of the day each price was read, not a permanent state. Each price on the product pages carries its own date. Source for the conversion: European Central Bank euro reference rate.
Finding two: the figures that do not agree
12 specifications across 6 models where two published sources give different numbers. 3 are cases of a manufacturer disagreeing with its own pages.
Dreame's own page says both 75 and 60 days
- Dreame L10s Ultra Gen 2Run time
240 vs 260 minutes across sources
- Dreame L10s Ultra Gen 2Navigation
Currys says RGB camera; Trusted Reviews says LiDAR only
- Dreame L50 Ultra AEClaimed suction
Dreame says 28,000Pa; Currys says 20,000Pa
100 days vs 75 days across Dreame pages
- Dreame L50 Ultra AENavigation
Described as two different systems across sources
- Dreame X50 Ultra CompleteRun time
Dreame and T3 say 220 min; Currys and TechRadar say 200 min
- Dreame X60 Pro Ultra CompleteSelf-empty interval
Dreame says 100 days; Currys says 75 days
eufy product page says 58dB; eufy comparison page says 56dB
- Roborock Saros 20Claimed suction
Roborock UK says 36,000Pa; TechRadar says 35,000Pa
- Roborock Saros 20Battery
Roborock UK says 6,400mAh; Expert Reviews says 6,300mAh
- Roborock Saros 20Threshold clearance
Roborock UK says 8.8cm double-layer (4.5+4.3); TechRadar says 4.5+4cm
And 2 prices the seller contradicts
iRobot's own UK site published two different prices for this model on 18 September 2026, with the colours swapped between them. Its Roomba listing page showed black at £129.00 (reduced from £229.00) and white at £299.00. Its products page showed white at £129.00 (reduced from £299.00) and black at £229.00. The seller contradicts itself about which version costs what, so we publish no price for either.
Our own earlier record carried £1,149 for this model, sourced to DJI's UK store. That page returned a 404 when re-checked on 19 September 2026, leaving the figure with no live source, so it has been removed rather than left standing.
Each of these appears as a blank cell on the model’s page. We would rather show nothing than pick the number that flatters the product, and neither figure is safe to quote as fact.
Finding three: 24 shop links that did not lead to the product
We researched these listings, could not confirm the page sells the exact model, and left them out. The worst case resolved to a Pilates machine.
All 24 were withheld for the same reason: we could not confirm the page sells that exact model. Two of the Currys links recorded during the research resolved to unrelated products — a Pilates machine and a Microsoft 365 bundle — which is what prompted us to check every retailer link by opening it.
Some retailers resolve a product page by numeric code and ignore the rest of the address, so a link that looks right can quietly serve something else. That is a problem for any site that lists prices, including this one — which is why a link only appears here once it has been opened and checked.
How to check this yourself
Nothing here depends on trusting us. Every row names its sources, and the links go to the manufacturers and reviewers rather than to us.
- Open the model's page on this site and find the blank cell. The same row appears above with both disputed figures and who published each one.
- Open the manufacturer's own product page in the country named. Prices on this site carry the date they were read, so compare like for like.
- Where a reviewer is cited, the review is linked on the model's page. We link tests rather than summarising them, so you can see the figure in its original context.
- If you find a figure we have wrong, or one that has since been corrected, tell us and we will change it and say that we did.
Why publish this at all
Doesn’t this make your own data look unreliable?
It makes the gaps visible, which is the point. Any site can print a full specification table by choosing one number whenever sources disagree. We would rather the blank cells be explained than quietly filled.
Are you accusing manufacturers of lying?
No. Most of these look like pages updated at different times, or a figure measured differently by a reviewer. The point is not motive, it is that a shopper comparing two robots is comparing numbers that do not always mean the same thing.
Will you take a price gap down if a manufacturer objects?
We will correct anything shown to be wrong, and we date every price so a stale one can be spotted. We will not remove an accurate figure because it is unflattering.
Does commission affect any of this?
No links on this site are paid links today. If any become paid links they will be labelled, and commission will not change what we publish or the order of any list. That rule is the reason this page exists.