4,752 Counterfeit Sales, None of Them Counted

TL;DR

• In Fendi v Rolo Fashion (IPEC, 9 July 2026), four LVMH houses won £213,000 against an influencer who sold counterfeit luxury goods for six years.
• Nobody could say how many items were sold. The court reached 4,752 by dividing £51,551.14 of deposits from one bank account by a £110 average price and extrapolating across 72 months.
• LVMH argued the goods were high-quality 'superfakes' and therefore displaced genuine sales. The judge rejected that on price: at under 15% of retail, 'it would take a naïve consumer to believe that products sold at those price levels are genuine.'
• Every question that decided the case — volume, suppliers, prices, dates — was a fact about transactions, and the only record of them sat with the infringer.

In August 2024, four LVMH houses froze the assets of an Essex influencer and got a court order for her bank statements. Sixteen months later they had a judgment in default, an admission of liability, and no idea how many counterfeit handbags she had sold.

The number the court eventually used was 4,752. It was not counted. It was produced by taking £51,551.14 of deposits from one Monzo account over seven months in 2024, dividing by £110, and multiplying the result across six years.

Fendi Italia Srl & Ors v Rolo Fashion Ltd & Anor was decided in the Intellectual Property Enterprise Court on 9 July 2026. Fendi, Loewe, Christian Dior and Celine, with LVMH itself as fifth claimant, had already won on liability in January 2025 when Georgia Aldridge and her company Rolo Fashion did not defend. Aldridge runs a social media marketing business and has more than 32,000 Instagram followers. The counterfeits were dropshipped, some of them from AliExpress, and sold in part through a dedicated WhatsApp group. All that remained was to decide what it was worth.

What the disclosure actually produced

The default judgment ordered the defendants to disclose their suppliers, the number of infringing goods received and the prices paid, who they sold to, and how many items they sold at what price. Three suppliers were named in the order: AliExpress.com, DHgate.com, and an entity called Xu Qiu, which the claimants had found only because a frozen PayPal account revealed 55 payments to a "Xu Qiu Ping."

What arrived, six weeks later, was an affidavit with AliExpress transaction records "but nothing else." Nothing on DHgate. When the claimants pressed with Part 18 requests, the defendants sent back an export of the WhatsApp group covering April 2023 to October 2024, in plain text, images excluded, retrieved from a spare device because the phone had been stolen.

So the brand-protection team built its estimate out of what it had. Nicolas Lambert, LVMH's Head of Online Brand Protection, counted 1,311 members in the WhatsApp group and assumed each one had bought at least one item. Over the 17 months of the transcript that gave 77.12 sales a month, which he extended across the 72-month claim period to reach 5,552 sales. The judge would not take it: "No basis has been provided for his assumption that each member … bought one product from the defendants."

Aldridge's counter-method was to use her bank statements. She had disclosed accounts at Halifax, PayPal and Monzo for January to August 2024, and she told the court these "provide a complete and accurate basis for assessing the number of sales of infringing products in the 72 months." She then used only the Monzo account, without explaining why, and exhibited only that one. It showed £51,551.14 in and £55,002.00 out.

HHJ Hacon accepted the shape of her method and corrected one input. She had divided her receipts by £751.67, the claimants' average retail price, which produced 69 sales. The relevant figure was what she charged, not what Fendi charged. "If £51,551.14 is divided by £110, the number of assumed sales would be just over 468, or around 66 sales per month." Multiplied by 72 months: 4,752.

That is the entire evidentiary basis for the volume of a six-year counterfeiting operation. One bank account, seven months, one division.

The superfake argument lost on price

Lambert's second argument was the one most brand-protection teams would make. The goods were not crude fakes. At least some were what online communities call superfakes or dupes, what the trade itself advertises as "1 to 1" or "mirror quality." High fidelity, close attention to detail, therefore buyers were deceived, therefore each counterfeit sale displaced a real one.

The judge was not convinced, and his reason had nothing to do with the objects. Lambert's own table put the defendants' prices "consistently and markedly lower, in one instance less than 5% of the third claimant's price," averaging a little under 15% of genuine retail. "It would take a naïve consumer to believe that products sold at those price levels are genuine." The messages from buyers, he added, did not read as people who thought they had bought a Dior. They read as people pleased with "very good imitations at prices which are much lower than those charged by the claimants."

The craftsmanship of the fake was argued at length and decided nothing. The price point settled it in a paragraph.

That finding then drove the whole award. If buyers knew, most sales displaced nothing. Hacon assumed a substitution rate of 15% — unrelated to the price ratio that happens to sit near the same number — and reached it by splitting the difference between the parties' silence, since the defendants' "probable failure to give proper disclosure" meant they got no presumptions in their favour, while nothing supported a high rate either. That gave around 713 lost sales at roughly £280 profit each: £199,640, rounded up to £200,000.

The other 4,039 sales still infringed, but had cost the claimants no sale at all. They were compensated on the user principle, a hypothetical licence the brands would never have granted. With no evidence to price it, the judge picked "a bare minimum" of 3% of the defendants' selling price. That is £13,328.70, rounded down to £13,000.

The reputational claim, the one head of damage that would have captured the brands' actual grievance, failed outright as "speculative and lacking in support." Buyers had not thought they were dealing with Fendi. They understood perfectly that they were in "a market which is deliberately run in parallel with the market for the claimants' goods."

£213,000, for six years, four houses and 4,752 assumed items.

What the court kept asking about

Read the judgment as an evidence problem rather than a trade mark one and something stands out. Almost nothing the court needed to know was a question about a handbag.

It needed the count. It needed the suppliers, which is why the order named Xu Qiu, DHgate and AliExpress specifically. It needed the buyers, the prices charged, the dates, the accounts money moved through. Every one of those is a fact about a transaction, not a fact about an object, and every one of them could have been answered from a complete record of what moved where. The physical fidelity of the goods was the one thing both sides genuinely contested, and it turned out to be the one thing that did not matter.

None of that record existed anywhere except inside the defendant's own bookkeeping, which is exactly where an infringer has every incentive to keep it thin. A stolen phone removed the images. One of three disclosed accounts got exhibited. DHgate simply never came up again. The court was not lied to so much as starved, and the arithmetic it fell back on — receipts over average price, times months — is what starvation looks like when a judgment still has to produce a number.

It is worth noticing which direction the gap runs. This was not a failure to tell a fake from a real one. Everyone involved, including the buyers, knew what these were. It was a failure to see the shape of a distribution network that ran for six years across at least three sourcing platforms, several payment rails and a WhatsApp group of 1,311 people, and that surfaced in court as one bank statement and a division. The question the record could not answer was never whether an item was genuine. It was how many, from whom, to where — and nobody, on either side of the case, was in a position to ask it.

References

• Fendi Italia Srl & Ors v Rolo Fashion Ltd & Anor [2026] EWHC 1703 (IPEC), HHJ Hacon, 9 July 2026: https://caselaw.nationalarchives.gov.uk/ewhc/ipec/2026/1703
• TheIndustry.fashion, "Influencer ordered to pay £213,000 over side-hustle selling counterfeit luxury fashion": https://www.theindustry.fashion/influencer-ordered-to-pay-213000-over-side-hustle-selling-counterfeit-luxury-fashion/
• Lewis Silkin, "An expensive side hustle: social media influencer ordered to pay £213,000 for trade mark damages", 10 August 2026: https://www.lewissilkin.com/insights/2026/08/10/an-expensive-side-hustle-social-media-influencer-ordered-to-pay-213-000-for-tra-102ngsx
• IP Twins, "Fendi v Rolo Fashion: Assessing Damages When Counterfeit Sales Do Not Translate into Lost Sales", 19 August 2026: https://iptwins.com/2026/08/19/fendi-v-rolo-fashion-assessing-damages-when-counterfeit-sales-do-not-translate-into-lost-sales/
• Europol and EUIPO, "Intellectual Property Crime and its Link to Other Serious Crimes: Focus on Poly-Criminality", June 2020: https://www.europol.europa.eu/publications-events/publications/intellectual-property-crime-and-its-link-to-other-serious-crimes-focus-poly-criminality

Image: "HM Courts & Tribunal Service Rolls Building -01" by Roger Green, licensed under CC BY-SA 4.0, via Wikimedia Commons. The Rolls Building houses the Business and Property Courts of England and Wales, including the Intellectual Property Enterprise Court.

Source

• Find Case Law, The National Archives: Fendi Italia Srl & Ors v Rolo Fashion Ltd & Anor [2026] EWHC 1703 (IPEC)