Deal Watch: The Cliff, and the Margin That Wasn't There

A UK-based LED beauty brand crossed my desk this spring. Dropshipped skincare and red light therapy devices on Shopify, founded in 2024, financials verified through connected Shopify and Google Analytics. The numbers are real. The seller just built the headline out of the eight months that worked and left off the two costs that make it a business.

The headline, as the marketplace frames it: a calendar 2025 with £412,554.97 in total sales across 5,393 orders, an average order value around £77, and a profit and loss statement showing £394,903 of seller's discretionary earnings on £433,409 of gross revenue. That is a 91% margin. The confidential memorandum that circulated in March softened it to a 49% margin on a £349,125 trailing-twelve figure, which is a quieter lie about the same business.

Now the price. The original ask was $81,440. It's been cut 35% to $53,319, about £39,282 today. The March memorandum pinned it lower still, at £30,000. Three documents, three asking prices, and a listing that has 1,222 views, 42 watchers, and exactly one comment with no sale. Before you read a single number, the market has told you the number won't hold.

The verified data has two things to add that the listing does not. The first is when the revenue happened. The second is whether the profit was ever real.

Here is the verified monthly gross revenue for 2025, straight off the connected Shopify export:

January £74,341. February £43,006. March £38,732. April £52,138. May £56,101. June £53,496. July £48,530. August £28,252. September £30,934. October £2,454. November £2,938. December £2,487.

Read that line again at the September-to-October seam. Revenue fell 92% in a single month and never came back. The fourth quarter produced £7,879 of revenue combined. January alone out-earned the entire fourth quarter more than nine times over. The first half of the year did £318K; the second half did £116K, and almost all of that was front-loaded into July.

Shopify's payout ledger, a second independent export, confirms it to the pound. Actual cash settled to the bank was £28,390 in September, then £2,713 in October, £2,579 in November, £2,063 in December. Into the new year: £1,450 in January 2026, then £211 in February. Two separate exports, revenue booked and cash received, agree exactly. The store didn't soften. It stopped in October, and it has stayed stopped for six straight months.

There's a subtler tell in the annual totals. 2024 did roughly £430K, 2025 did £412K, and a buyer glancing at year-over-year sees a stable, barely-declining brand. The monthly cash says otherwise. The 2024 payouts ramped hard into a December 2024 peak of £82K; the 2025 payouts started high and fell all year. Stack the two and the truth is that this brand peaked around the turn of 2024 into 2025 and has been declining for fourteen months, ending in a dead stop. The flat annual comparison is an accident of one year being back-loaded and the next being front-loaded. The momentum at the handover is straight down.

Now the profit. The P&L shows £394,903 of earnings on £433,409 of revenue, a 91% margin, and it reaches that number by leaving out the two largest costs any paid-acquisition dropship brand carries.

Look at the cost of goods line. It is £833.33. Every month. It is £833.33 in January when the store does £74,341, and it is £833.33 in October when the store does £2,454. A cost of goods sold that is identical whether you sell £74K or £2K of product is not a cost of goods sold. It's a placeholder. For the full year it totals £10,000, which is 2.3% of revenue, on physical LED devices that have to be bought before they're shipped.

Then look for the advertising line. There isn't one. The listing states plainly that the brand is acquired through paid advertising, and the single largest expense in that model, the ad spend that buys the traffic, appears nowhere on the statement. The verified analytics show 218,280 sessions for the year, up 18% over 2024, against a conversion rate of 1.98% that fell 14%. That is the signature of paid traffic being bought harder and converting worse, and none of its cost is on the P&L.

So price the business honestly. Book a realistic product cost for dropshipped devices, call it 35% of revenue, and a paid-acquisition spend consistent with the ad-driven model the seller describes, and the £395K of "profit" collapses toward a fraction of itself, plausibly into the 15% to 30% margin range in the months that worked at all. The exact figure is for diligence to find. The point is that the 91% on the page, and the 49% in the memorandum, are both manufactured by omission. You are not being shown a high-margin brand. You are being shown revenue with the costs deleted.

Apply that to what the store does now. The fourth-quarter run rate annualizes to about £31K of revenue. Strip in real product cost and the minimum viable spend to keep a Shopify store alive, and the trailing run-rate profit is near zero, quite possibly negative. At zero profit the multiple isn't 0.2x or 49x. It's undefined. You aren't buying an earnings stream. You're buying assets, and you should price them as assets.

This is the distinction that matters, and it's what separates this deal from an ordinary declining brand.

A gradual decline, several quarters of softening, AOV slipping, CPMs creeping up, is a story a buyer can underwrite. You can name the cause, model a recovery, and decide whether you can run the play better than the seller did. That's a real deal with a real discount.

A vertical drop from £30,934 to £2,454 in a single month, held at the floor for six months, is a different animal. That shape doesn't come from tired creativity. It comes from a switch being thrown: an ad account ban, a payment-processor or Shopify suspension, a dropship-fulfillment relationship that ended, a chargeback or compliance event, or an operator who simply stopped spending because the unit economics finally inverted. The package discloses none of it. It doesn't acknowledge the cliff exists. The buyer is being asked to underwrite a recovery without being told what broke.

That's the bull case: a verified history and a complete handover at a low headline price. The bear case lives in four places.

The decline has no disclosed cause. Revenue collapsed 92% in October and has stayed on the floor for six straight months. The package doesn't mention it, let alone explain it. The single most important fact about this business is the one fact nobody wrote down.

You're buying into a "growing market" this brand just exited. The listing leans on structural demand for red light therapy across EMEA, the GCC, and Asia-Pacific. Grant the premise in full. It makes the story worse, not better. In a rising category, with that tailwind, the brand still fell off a cliff. The market isn't the problem. The problem is the brand, the funnel, or the operator, which are the three things you'd be buying.

The market has already voted. Forty-two sophisticated buyers put this on a watchlist, looked at the same verified exports you're looking at, and declined to bid, even after a 35% price cut. A price that has been reduced and still hasn't cleared is the market telling you where value actually sits.

The moat is rented. It's a dropship model selling LED devices that hundreds of other Shopify stores and Amazon sellers offer at lower prices. No proprietary formulation, no patent, no exclusive supply. The "supplier contracts" are dropship arrangements available to anyone who asks. The only durable assets here are the aged domain, the brand mark, the creative library, and the email list.

The seller isn't being careless. He's a solo operator with a thin marketplace track record, not an institutional portfolio recycling capital, and he has already marked the price down 35% without a taker. That's a person trying to recover something from an asset that stopped working, in a process that hasn't found its floor.

A brand that peaked fourteen months ago and then cratered is worth its salvage value, not a multiple of an earnings stream that was never quite real and no longer prints at all. The falling price isn't a discount being handed to you. It's a discovery process that hasn't bottomed.

The buyer's job is not to assume the seller is wrong about the asset. The buyer's job is to know which asset they're actually being sold.

There's a buyer for this. It is emphatically not the person looking for a turnkey, cash-flowing brand. That person reads the £395K profit line, signs the LOI, and gets the verified monthly P&L two weeks later with a look on their face.

The buyer for this is someone for whom the price is an asset purchase, not a business purchase. A parts bin: a two-year-old domain in a hot category, an email list, a tested creative library, a supplier introduction, and a built Shopify store. If you already run a DTC stack, your own ad accounts, your own email infrastructure, your own creative pipeline, and you can relaunch those parts into operations you control and cost honestly, then a low number for the bundle might pencil. If you're buying the cover page, you're buying a ghost.

If that's you, the playbook is short:

The package sells a 91% margin and a brand on a growth-market tailwind. The verified exports sell the truth: a brand that peaked at the turn of the year, ran on borrowed costs, fell 92% in a single month, and has been dead for six. Both stories live in the same file. The buyer's only job is to read the one the seller hoped you'd skip.

Subscribe for the next teardown. I read these so you don't sign blind. Forward this to anyone about to put an LOI on a cover page. And if you're looking at a deal of your own, reply to this email and walk me through it.