Black Friday 2026: the made-up brands shoppers click on

By Josh Pigford · Published November 13, 2026 · Data from October 14 to November 13, 2026

In the month leading into Black Friday, Knockoff users opened Amazon.com product pages from 10,470 different brand names that Knockoff’s detector rates as likely made up: mostly strings of letters that spell nothing, with no reputation behind them. That count comes from one browser extension’s users, on one storefront, in 30 days.

US Amazon.com only. Products a Knockoff user opened from October 14 to November 13, 2026. Figures read on November 13, 2026 and will not refresh.

What “made up” means here

Amazon search results are full of brand names nobody has heard of, built to clear a trademark office rather than to win customers. They are legal. They are not counterfeits. The catch is that a name like that can be dropped and replaced the moment its reviews turn bad, so it never has a reputation to protect. The longer explainer is Why is Amazon full of brands like these?

Knockoff’s browser extension rates every brand name it sees. This report counts a name only when the extension gives it its strongest rating, the one it acts on even at its most relaxed setting. Names it is unsure about are left out.

Where they turn up

We sorted the products into ten common gift categories by words in the product title. Kitchen and car accessories had the most distinct made-up names, with more than 500 each. Christmas decor came third.

Methodology

The data. When a Knockoff user opens an Amazon product page, the extension sends the product’s ID, title, brand, and seller to Knockoff (users can turn this off in settings; see privacy). The product record keeps no user identity. This report uses every US Amazon.com product with a brand name that was opened at least once from October 14 to November 13, 2026: 999,076 products from 193,393 brand names.

The rating. Each brand name went through the extension’s own brand detector, the same code that runs in shoppers’ browsers. It checks a list of 9,299 known real brands first, so a known brand is never counted. A name counts here only if it is on Knockoff’s hand-picked junk list or scores high enough on the detector’s name tests (long runs of consonants, all capital letters, odd mid-word capitals, spelling that reads as no language the detector knows). 10,470 names passed that bar.

Spot check. We read a random sample of 80 of those names by hand. 66 were plainly random strings of letters. The other 14 were short initials or mixed names that could belong to a real small company. That is why this page calls the whole count likely made up, not proven.

Categories. A product lands in a category when its title contains that category’s keywords, checked in a fixed order. It is a rough sort, not Amazon’s own category tree.

What these numbers do not claim

• They name nobody. Everything here is a count. No brand or seller is identified, and none should be inferred.
• This is not a share of search results. The data covers product pages people opened, not everything they saw in search. Products from likely made-up brands were 2.9% of the products opened (29,096 of 999,076). That share likely runs low: by default the extension dims these listings in search, so its users click on fewer of them than other shoppers would.
• Made up is not the same as dangerous. Plenty of these products work fine. The problem is that the name tells you nothing, and it can disappear with its reviews.
• Knockoff users are not all shoppers. They are people who installed a filter for exactly this, browsing the US store.
• It misses some, and catches a few it shouldn’t. The detector looks only at the name. A made-up brand with a pronounceable name gets past it, and the spot check above found short initials that could be real companies.

Cite this page

For journalists

The queries behind every figure are available on request. Write press@knockoff.co. Logos, screenshots, and boilerplate are in the press kit. For background on who files these trademarks, see the Pseudo-Brand Index.