Consumer Group Sues Amazon Over Excessive Thumb Scrolling After Sponsored Products Reportedly Take Over Search Results

Local Nation — A consumer advocacy group filed a class-action lawsuit against Amazon Friday, alleging that excessive sponsored-product placement has forced online shoppers to perform dangerous quantities of thumb scrolling merely to reach the products they were actually looking for.

The fictional complaint, American Consumers With Tired Thumbs v. Amazon, alleges that searching for an ordinary product increasingly requires users to scroll past sponsored listings, recommended listings, promoted brands, related products, deals, and several items that appear to have absolutely nothing to do with what the motherfucker typed into the search box. The complaint runs to 212 pages. The word “sponsored” appears in it 1,906 times, which the plaintiffs’ counsel describes as “a fair sample.”

The lead attorney opened the filing-day press conference with two sentences and one exhibit. The Division reproduces them in the order delivered.

“Our clients came to purchase batteries.”

He displayed a phone.

“They were not prepared for cardiovascular exercise.”

Our Research Division has obtained the complaint, its exhibits, the plaintiffs’ investigative search logs, the deposition of the lead plaintiff and the separately sworn statement of his mother, Amazon’s two corporate statements, the transcript of the company’s press briefing, the slide deck titled AMAZON CUSTOMER THUMB SAFETY REPORT, the Thumb Wellness at Amazon launch materials, the ergonomics working paper defining the Thumb Scroll Distance, the sponsored-label legibility test, the labor economists’ wage estimate, the consultants’ Prime Thumb+ proposal and its fine print, two withdrawn regulatory proposals, the settlement term sheet, and the courtroom recording of the plaintiffs’ closing argument. The Division presents its findings below and states its principal conclusion at the outset, because every document in the file is an application of it.

The lawsuit is being covered elsewhere as a frivolous complaint about a sore thumb. The Division’s file supports a narrower and more instructive reading. Every sponsored placement in the record is a transaction between two parties, the advertiser and the platform, and every one of them is settled in the currency of a third. The advertiser pays Amazon for position. Amazon delivers the position. The position has a physical cost — it is the distance between the top of the screen and the thing the customer asked for — and that cost is paid in full, at the moment of delivery, by the only party in the transaction who was not consulted and is not compensated: the shopper, by thumb. The product the customer searched for was always free to display. What is sold is the distance to it. The thumb is not the victim of the business model. It is the medium of exchange.

The Division’s reconciliation of the three accounts involved in a single sponsored placement is attached for the record:

The Division also notes, at the outset, that the lead plaintiff’s injury is real, that it is extremely minor, that his mother kissed it, and that it is nonetheless the first unit of this cost anyone in the history of online retail has been able to point at.

CLASSIFICATION: CONSUMER PROTECTION — SEARCH RESULT OCCLUSION AND THE DISPLACEMENT OF ADVERTISING COSTS ONTO THE HUMAN THUMB
DISTRIBUTION: Consumer Advocates, Platform Regulators, Ergonomics Researchers, Hand Surgeons, Mothers, Anyone Who Has Typed the Exact Name of a Product and Been Shown a Different One
PREPARED BY: The Externality Research Division
DATE: October 2026

Investigation Documents Severe Thumb Mileage

Before filing, the plaintiffs commissioned a field investigation. Researchers were instructed to search for a single common item and to record every element displayed before the item appeared. The Division has obtained the session log and reproduces it without abridgment, because the abridgment is the complaint.

They entered the term. Results appeared.

The lead researcher lowered the phone.

“Where’s the fucking cable?”

A research assistant, reviewing the same screen, answered with what the Division regards as the strongest single sentence in Amazon’s defense.

“Those are cables.”

The researcher’s reply is reproduced in full.

“Where are the cables Amazon would’ve shown me if nobody had paid them to show me different cables first?”

Nobody immediately knew how to answer.

The Division has examined the exchange at length and finds both parties correct, which is the problem. The assistant is right: every sponsored listing in the session was a USB-C cable, and several were good ones. No court will find that a shopper who asked for a cable was injured by being shown a cable. The researcher is also right, and his question names the only product in the session that Amazon could not display. The list he was asking for — the ranking the search engine would have produced on relevance alone, before position was sold — does not exist anywhere on the page. It is not hidden below the fold. It has been dissolved into the results, its first several places auctioned and the rest renumbered, so that the shopper is no longer able to see where the paid list ends and the earned list begins. The Division files this as the file’s founding observation: the counterfactual ranking is the one item the world’s largest store does not stock.

Session 001 required 23 thumb movements to reach the first unsponsored cable matching the query. The plaintiffs’ investigators ran 400 further sessions across ordinary household queries. The median was 21. The longest, for a replacement remote control for a television the researcher no longer owned, was abandoned at 96 thumb movements when the researcher, in the words of the log, “stopped believing the remote existed.”

Consumer Suffers Alleged Boo-Boo

The lawsuit’s most emotionally devastating testimony came from one customer who said prolonged scrolling caused a minor injury. The Division reproduces the reporter’s interview as recorded.

The reporter asked him to describe what happened.

“I was trying to find the regular products.”

“Okay.”

“But there were sponsored ones.”

“Then what?”

“I kept scrolling.”

“For how long?”

“A lot.”

The customer raised his thumb. A tiny irritated spot was visible. The room became silent.

“I got a boo-boo.”

The reporter asked whether it was serious. The customer shook his head.

“It kinda hurt.”

“Did you seek medical attention?”

“No.”

“What did you do?”

The customer looked down.

“I CALLED MY MOM TO KISS IT.”

Several consumer advocates reportedly had to leave the room.

The Division has reviewed the deposition from which this interview was later reconstructed under oath and wishes to record, before the file proceeds, that it takes the injury seriously in exactly the measure the injury deserves, which is very little, and that this is the point. The lead plaintiff’s boo-boo is not the harm in this case. It is the receipt. For as long as sponsored placement has existed, its cost to the shopper has been paid in units too small to notice: three seconds here, eleven there, a worse cable bought because it appeared first, a price a few cents higher because the seller’s advertising budget had to come from somewhere. None of these costs leaves a mark. The thumb, after an afternoon of looking for batteries, did. It is the only place in the entire consumer economy where the cost of an advertisement has ever become visible on the body of the person who paid it, and it showed up, as small costs eventually do when they are charged to the same account long enough, as a tiny red spot.

Mother Confirms Treatment

Investigators interviewed the man’s mother. Her statement was taken by the plaintiffs’ attorney in the presence of a court reporter and is reproduced without alteration.

“Did your son contact you regarding an Amazon-related injury?”

“He said his thumb hurt.”

“And what treatment did you provide?”

“I kissed it.”

“Did that resolve the condition?”

“Apparently.”

The attorney turned toward the jury.

“LADIES AND GENTLEMEN, AMAZON FORCED THIS MAN’S MOTHER BACK INTO ACTIVE MEDICAL SERVICE.”

Amazon’s fictional counsel objected.

“She kissed his thumb.”

The plaintiffs’ attorney did not sit down.

“BECAUSE OF YOUR CLIENT.”

The Division records the exchange as the case’s first itemized pass-through. The cost of a sponsored placement was charged to the shopper’s thumb. The shopper, lacking any channel through which to bill it back to the platform or the advertiser, did what every party at the end of a cost chain does, and passed it to the next person who would accept it without an invoice. That person was his mother. The Division notes that unpaid care is the oldest and largest externality on file in this publication’s archive — larger than carbon, older than money — and that it has now been documented, for the first time, absorbing the downstream cost of a keyword auction. The Division is not aware of any line in Amazon’s advertising revenue that is offset by maternal labor. It is also not aware of any reason to think there isn’t one.

Amazon Issues Corporate Response

Amazon’s fictional corporate communications department responded to the lawsuit Friday afternoon with a carefully worded statement insisting that the company takes customer thumb health extremely seriously.

The statement continued:

The injured customer read the statement in the presence of reporters and an Amazon spokesperson.

“So it’s my fault?”

“That’s not what the statement says.”

“You told me to take a break.”

“We encouraged appropriate ergonomic behavior.”

“I WAS TRYING TO BUY FUCKING BATTERIES.”

The Division has performed a line-by-line review of the statement and certifies that every sentence in it is true. Sponsored products can help customers discover brands. The company does dispute the allegations. Customer safety may well be a priority. And a break would, in fact, relieve the thumb. The Division draws attention only to the direction in which the final sentence points. It identifies a harm produced by the design of the page and proposes a remedy to be performed by the person on the other side of the glass. This is the standard form of every corporate response in the externality literature, and the Division has a name for it in its internal style guide: abatement by the injured. The factory recommends that residents close their windows. The platform recommends that the shopper rest his thumb. In neither case does the recommendation contemplate any change to the thing producing the smoke, and in both cases the recommendation is, on its own terms, perfectly sound medical advice.

The Division also notes the precise wording of the denial — clinically significant thumb boo-boos — and observes that it concedes the boo-boos, disputing only their clinical significance, a position the plaintiffs’ own medical expert will later be unable to contest.

Amazon Says Sponsored Products Improve Discovery

At a corporate press briefing, executives stressed that advertisements should not be understood as obstacles between customers and search results. The Division has obtained the transcript.

“Sponsored products are themselves products.”

“Sponsored products?”

“Correct.”

“That advertisers paid to place prominently?”

“They are useful discovery opportunities.”

“Ahead of the thing the customer searched for?”

“Potentially alongside relevant shopping experiences.”

“Is that corporate for yes?”

The executive took a sip of water.

“We reject the premise of the question.”

A second reporter followed.

“How many thumb movements should it take to find the exact product I typed?”

“Shopping journeys differ by customer.”

“Three?”

“It depends.”

“Ten?”

“Again—”

“Thirty?”

“We don’t characterize customer discovery in those terms.”

The reporter held up his phone.

“MY THUMB CHARACTERIZES IT IN THOSE TERMS.”

The Division certifies the executive’s opening sentence — sponsored products are themselves products — as the most accurate statement in the company’s defense and its entire content. It is a tautology, and like all good tautologies it cannot be refuted, only weighed. Every sponsored product is a product. The question the briefing was convened to avoid is not whether the sponsored cable is a cable but why it is first, and the executive’s four successive substitutions for the word “first” — prominently, discovery opportunities, alongside, relevant shopping experiences — were each reviewed by the Division’s copy desk, which reports that none of them means first and all of them describe it.

The Division also records the second reporter’s question as the first attempt in the file to put a number on the shopper’s side of the transaction, and the executive’s refusal to “characterize customer discovery in those terms” as a complete description of how the transaction has been priced to date. On one side of every sponsored placement there is a bid, an impression count, a click-through rate, a cost-per-click, and an attributed sales figure, each measured to the fourth decimal place. On the other side there is a thumb, and the company does not characterize it in those terms. The reporter’s thumb did. The Division notes that the only metric the platform does not track is the only one the customer can feel.

Amazon Unveils Internal Thumb Safety Data

The company presented research showing that the overwhelming majority of customers complete shopping sessions without reporting thumb injuries. The Division reproduces the principal slide.

A reporter raised her hand.

“How do customers report a boo-boo?”

“Through customer service.”

“Where?”

“Customers can navigate to Help.”

“Then?”

“Select the appropriate issue.”

“Then?”

“Follow the prompts.”

“How many screens?”

The executive stopped. The consumer attorney stood in the back of the room.

“DON’T SAY IT.”

“It varies.”

“GODDAMMIT.”

The Division has audited the 99.999% figure and finds the arithmetic sound and the denominator disqualifying. The statistic measures searches completed without a reported injury, and the only channel for reporting one is a sequence of screens navigated by thumb. An injury to the thumb therefore reduces the injured party’s capacity to report it, in proportion to its severity, so that the most seriously affected customers are the least likely to appear in the numerator. The Division recognizes the structure. It is the reporting system of every workplace that requires an injured worker to fill out the form with the injured hand, and it produces, everywhere it is used, a safety record of extraordinary quality.

The Division has computed the figure on the slide in absolute terms. Applied to the plaintiffs’ estimate of annual product searches on the platform, a reported-injury rate of 0.001% is several million boo-boos a year. The company did not present this conversion, and the Division does not suggest that it should have, the slide’s purpose being to describe the boo-boos that made it through.

Injured Customer Attempts to Report Injury

The plaintiff agreed to test Amazon’s customer-service process in the presence of reporters. The Division has reconstructed the session from the video.

He opened the app.

The customer stopped. A reporter asked him what was wrong. He held up his hand.

“OTHER THUMB.”

The plaintiffs immediately amended the complaint.

The Division files the second injury as the case’s controlled experiment, and the cleanest result in the file. The customer had switched hands to protect the first thumb. He was not searching for a product and was shown no sponsored listings. He was attempting to report a harm through the channel the company had designated for the purpose, and the channel produced, in eleven inputs, a second instance of the harm it exists to record. The Division notes that no part of the customer-service flow is advertising and that it nonetheless reproduced the injury exactly, which suggests that the injury was never caused by the advertisements as such. It is caused by distance — by any interface in which the thing the customer wants has been placed some number of thumb movements below the place the customer starts — and the sponsored listing is merely the most profitable way of producing it.

Amazon Denies Creation of Secondary Boo-Boo

Corporate communications released another statement.

The consumer attorney read it aloud at the next briefing.

“Alleged?”

“We have not independently verified the boo-boo.”

The man’s mother entered the press room.

“I verified it.”

Every camera turned.

“Ma’am, are you a licensed medical professional?”

“No.”

“Then—”

She held up her son’s hand.

“I kissed this one too.”

The room erupted. The plaintiffs’ attorney pointed toward her.

“TWO THUMBS. ONE MOM. HOW MANY MORE FAMILIES MUST SUFFER?”

“This press conference has become ridiculous.”

“YOUR SEARCH RESULTS STARTED IT.”

The Division has reviewed the question of verification with some care, because it is the question on which every externality case eventually turns. Amazon’s position — that the boo-boo had not been independently verified — is procedurally correct. No physician examined either thumb. No imaging was performed. The only witness to both injuries is a woman with no medical license and an obvious interest in the outcome. But the Division notes that she is also the only party in the dispute who has handled the evidence, that her treatment was administered within hours of each injury and was in both cases effective, and that the company’s demand for independent verification of a cost it produced, using instruments it does not provide, from a professional class it does not pay, is the externality’s defense in its purest form. A cost that leaves no records was not incurred. A record kept by the person who bore it is not independent. Between those two rules lies every harm in this publication’s archive.

The Division further notes that the spokesperson’s closing assessment — that the press conference had become ridiculous — is accurate, and that the plaintiffs’ attorney’s attribution of causation is also accurate, and that the Division can find no way to dispute either.

Company Announces Thumb Wellness Initiative

Within forty-eight hours, Amazon announced a new program. The Division reproduces the launch materials.

A consumer reviewed the fourth component.

“Alternate fingers?”

“Yes.”

The consumer began scrolling with his index finger. Ten minutes later:

“Now this motherfucker hurts.”

Amazon’s legal department reportedly asked engineering to stop releasing features.

The Division has reviewed each of the four components and finds that none of them alters the distance between the top of the page and the product. Three of the four address the shopper’s capacity to bear the distance, and the fourth, shopping accessibility research, has not yet produced a finding. The finger-alternation reminder, which the Division regards as the initiative’s only operative clause, is recognizable to anyone trained in risk management: it is diversification. The total cost of the sponsored placements is unchanged. It has merely been spread across a larger portfolio of digits, reducing exposure on any single finger while increasing the number of fingers exposed, and the consumer’s review of it after ten minutes is, in the Division’s assessment, the correct actuarial result. A cost that has been diversified has not been reduced. It has been given more places to hurt.

The Division also records the legal department’s request to engineering as the first instance in the file of anyone at the company moving to reduce the rate at which the problem was being produced.

Experts Establish Thumb Scroll Distance

Ergonomics researchers retained by the plaintiffs developed a new unit of measurement, which the Division reproduces with its definition.

The working paper then classifies search sessions by the distance required to reach the first result that matches the query on relevance alone.

The Division welcomes the unit, which is the first instrument in the history of e-commerce to measure the shopper’s side of the sponsored-placement transaction in the shopper’s own currency. It notes that the working paper’s final category is not a measurement but a prayer, and that the researchers explained in a footnote that the search in question did not terminate during the observation window and that its distance was therefore classified as undefined. The Division considers this generous. A search that does not terminate has a distance; it is simply not finite, and the researchers declined to say so in a document that would be read by a jury.

The Division’s chart of the working paper’s values is reproduced below.

The Division notes that the distance between the first two bars — from a historical baseline of four to seven thumb movements to a modern figure of thirty-one — represents roughly twenty-five TSD per search that were not required by the product, the query, the catalog, or the technology, all of which were the same at both ends of the chart. They are the inventory. Somewhere in the platform’s accounts, those twenty-five scrolls appear as a line of revenue. In the shopper’s accounts, they appear nowhere at all, which is why, until this lawsuit, they had never been counted.

Sponsored Label Found Suspiciously Small

Consumer advocates enlarged screenshots during the hearing and asked the jury to identify which results were advertisements. The Division reproduces the exchange from the transcript.

“Can anyone identify which result is sponsored?”

“That one?”

“No.”

“That one?”

“Sponsored.”

“What about that one?”

“Also sponsored.”

“That one?”

“Regular.”

“They look almost identical.”

“EXACTLY.”

A graphic designer was then asked to locate the sponsored disclosure on an actual phone, held at a measured distance. The Division reproduces the test record.

The courtroom reportedly gasped as though someone had discovered a sniper.

The Division has reviewed the test and draws attention to the final distance. A person reading a phone holds it, by most measures, somewhere between twelve and sixteen inches from the eye. The disclosure became legible at eighteen. It is therefore visible at almost exactly the distance at which the reader’s thumb is already on it — the label arriving, in effect, at the same moment as the listing it labels, too late to be avoided and early enough to have been disclosed. The Division does not suggest that the label’s size was chosen for this property. It notes only that a disclosure designed to be read before a decision would look different, and that a disclosure designed to have been present afterward would look like this.

The Division also records the jury’s error rate. Of four results, the jurors classified two correctly by guessing and two incorrectly by looking. The Division regards this as the correct experimental outcome for a label whose function is to be technically present: the closer the advertisement resembles the result, the more valuable the position, because a shopper who cannot tell the difference cannot discount it. The disclosure and the camouflage are, in this design, the same object.

Consumer Group Claims Search Has Become Advertising With Search Results Hidden Inside

The complaint argues that traditional search followed a simple architecture, which the Division reproduces as diagrammed in Exhibit 12.

Researchers claim the emerging architecture resembles the following, reproduced as diagrammed in Exhibit 13.

The plaintiffs have named the phenomenon:

Amazon disputes the characterization.

The Division has compared the two exhibits and observes that the thing is present in both. Nothing has been removed from the second architecture. The answer to the user’s question is still on the page, still correct, still for sale at the same price. What has changed is its address. In the first diagram the thing is one step from the question; in the second it is seven, and the six intervening steps are, in the plaintiffs’ accounting, the product. The Division notes that “occlusion” is an apt term borrowed from medicine and astronomy, where it describes an object that is still there but has something in front of it, and that in both disciplines the object doing the occluding is not considered to have improved the view.

The Division also draws attention to the final line of Exhibit 13. The architecture does not end with the thing. It continues past it. The researchers explained that this is because the platform does not regard the shopper’s arrival at the product as the end of the session but as a fresh impression opportunity, and that a shopper who has found what she was looking for is, from the auction’s point of view, at her most valuable.

Plaintiff Searches Exact Product Name

Attorneys demonstrated the problem in court by searching an exact hypothetical product. The Division reproduces the query and results as projected.

The plaintiff, operating the phone, addressed the screen.

“I TYPED THE FUCKING NAME.”

A researcher seated at counsel table answered.

“Keep scrolling.”

“WHY?”

“It’s down there.”

“YOU KNOW WHERE IT IS?”

“Yes.”

“THEN PUT THAT MOTHERFUCKER UP HERE.”

The Division files the researcher’s one-word answer — yes — as the most important admission in the case, and notes that it was made by the plaintiffs’ own witness, who meant it as encouragement. The exact-match query is the controlled condition of the entire dispute. In an ordinary search, a platform can argue that relevance is a judgment, that shoppers differ, and that a sponsored listing may be what the customer would have chosen had he known about it. A query for a specific model number of a specific replacement part from a specific manufacturer removes every one of those arguments. There is one right answer. The system knows what it is. The system knows where it is. Its position on the page is not a limitation of the search engine. It is a decision, made fresh for every shopper who types the name, and the decision is sold.

The plaintiff’s instruction — put that motherfucker up here — is, the Division notes, a complete specification of the remedy, requiring no new technology, no research program, and no additional engineering, the search engine having already performed the only difficult part of the job, which is finding the thing. The Division estimates the implementation cost of the instruction at zero and its revenue cost at four sponsored placements per exact-match query, and observes that the second figure is the entire reason the first has not been incurred.

Thumb Doctors Report Emerging Condition

A fictional medical association announced it was studying a new presentation. The Division reproduces the association’s preliminary clinical description.

The association’s physician testified for the plaintiffs. The Division reproduces the cross-examination’s final exchange.

“Clinically, most cases are extremely minor.”

“But they hurt?”

“Potentially a little.”

The attorney turned toward the cameras.

“YOU HEARD THE DOCTOR.”

“Please don’t use me like this.”

The Division has reviewed the symptom list and observes that its first three entries are physiological and its last three are social. The physician confirmed under oath that the social symptoms are the more reliable diagnostic indicators, because they are the only ones the patient reports without being asked. The Division further observes that the condition’s severity scale has a single upper bound, CALLING MOM, which corresponds not to any increase in tissue damage but to the moment at which the patient seeks a party willing to acknowledge it. The Division regards this as the correct clinical definition of a severe externality: not one that does great harm, but one that has finally found someone to tell.

The physician’s final request is noted. The Division sympathizes, and records that every expert witness in the history of litigation has made it, and that it has never once been granted.

Class Action Expands Rapidly

Within days, millions of consumers attempted to join. The Division reproduces the eligibility questionnaire as posted on the plaintiffs’ website.

Attorneys announced they may represent approximately:

The Division has reviewed the class definition and certifies it as the signature of an externality. A cost borne by a few people is a grievance, and they sue. A cost borne by everyone, in amounts too small for any one of them to notice, is a condition, and nobody sues, because no individual claim is worth the filing fee and no individual claimant is distinguishable from the weather. The questionnaire’s achievement is to convert the second kind of cost into the first: to find, in five yes-or-no questions, a way to describe a harm so universal that it had previously been indistinguishable from shopping. The Division notes that the questionnaire has a 100% qualification rate among respondents who reached the end of it, and that the plaintiffs’ firm has had to add a sixth page to its intake form for respondents who report that they could not reach the end of it because of their thumb.

Amazon Argues Sponsored Results Are Still Relevant Products

The company’s fictional attorneys noted that sponsored listings frequently remain related to a shopper’s query. The Division reproduces the colloquy between Amazon’s counsel and the lead plaintiff from the deposition transcript.

“That’s not the point.”

“What is the point?”

“I searched for something.”

“Yes.”

“You know what I searched for.”

“Yes.”

“You have it.”

“Yes.”

“You know where it is.”

“Yes.”

“But first I have to look at what somebody paid you to put in front of it.”

The transcript records a pause.

“NOW WE’RE FUCKING TALKING.”

The Division files the colloquy as the case’s complete theory of harm, delivered by the plaintiff under questioning designed to elicit something else. Amazon’s counsel conceded four facts in four syllables: that the platform knew the query, possessed the product, knew its location, and — implicitly, in the yes that never had to be said — had placed something in front of it for money. The plaintiff’s final sentence then identified the only element of the transaction that the company’s relevance defense does not reach. The sponsored product may be relevant. It may be good. The shopper may even buy it. None of this alters the fact that the shopper was made to look at it, in a sequence determined by payment, on the way to a thing the company could have shown him first. The objection is not to the product. It is to the toll.

The Division also notes that the line attributed to the transcript’s final speaker — now we’re fucking talking — is not marked in the record as belonging to either party, and that both sides’ counsel have since claimed it.

Thumb Workers Demand Compensation

Labor economists unexpectedly entered the dispute, filing an amicus brief arguing that every scroll represents uncompensated mechanical labor performed by the consumer. The Division reproduces the brief’s central calculation.

The lead plaintiff, attending the economists’ press conference, raised his hand.

“Where’s my money?”

“This isn’t an actual wage claim.”

“You shouldn’t have shown me the fucking number.”

The Division has checked the arithmetic and confirms it: 23 movements, 14 searches, twelve months, 3,864 movements, $38.64. The Division regards the one-cent rate as arbitrary and low, and notes that the economists chose it for exactly those reasons, so that no one could accuse them of inflating it.

The Division files the plaintiff’s response as the most important economic observation in the file, and the most general. A cost that has no number is a feeling. A cost that has a number is a debt. The economists did not discover anything about the shopper’s thumb that the shopper did not already know; he had been doing the scrolling. What they did was attach a figure to it, and the figure transformed the scroll from a minor irritation into an unpaid invoice, which the plaintiff immediately attempted to collect. The Division observes that this is why the platform does not “characterize customer discovery in those terms” — not because the terms are inaccurate, but because, once characterized, the cost has an owner, and the owner wants his thirty-eight dollars and sixty-four cents.

Amazon Prime Thumb Insurance Proposed

Consultants hired to solve the crisis immediately made everything worse. The Division has obtained their proposal, which is reproduced as presented.

An executive reviewed the slide.

“Absolutely not.”

“Consumers have demonstrated willingness to pay for convenience.”

“GET OUT.”

The consumer attorney was watching through the conference-room window.

“No, let him cook. I need Exhibit 94.”

The Division has reviewed the proposal and finds it the most candid document in the case, which is why the executive wanted it out of the room and why the attorney wanted it in the record. Every benefit on the list is the removal of a cost the platform is currently imposing. Reduced sponsored scrolling is the absence of something the platform adds. Priority organic results is the order the results would be in if it did not sell the order. The proposal does not offer the customer anything new. It offers to sell him back the search he had before, at $4.99 a month, and in doing so it prices the externality: the consultants have calculated, to the cent, what the platform believes it is worth to a customer not to be made to scroll. The Division notes that a company that can name a price for removing a cost has conceded that it is imposing one.

The fine print is then the proposal’s second, smaller confession, which is that even the paid version would not be free of the thing being paid to remove. The Division records the asterisk as the business model’s own assessment of its durability.

The Division has confirmed that the proposal was entered into evidence as Exhibit 94.

Government Considers Mandatory Scroll Breaks

Regulators proposed requiring apps to warn users after excessive thumb activity. Under the draft rule, after twenty consecutive sponsored-result scrolls, the following notice would be displayed:

A consumer testifying at the public hearing reviewed the draft.

“CONSIDER SHOWING ME THE FUCKING PRODUCT.”

Proposal withdrawn.

The Division notes that the regulators’ draft, like Amazon’s first statement, located the remedy in the person bearing the cost, and that the government arrived at this position independently of the company, which the Division regards as either a coincidence or a description of how regulation works. The Division also notes that the proposed warning would have appeared after twenty sponsored scrolls, which is to say it would have interrupted the shopper at approximately the point where the product was about to come into view, adding one more thing to the distance it was meant to warn him about.

Search Results Receive Nutrition-Style Label

A second proposal would require search pages to display a standardized disclosure modeled on the food label. The Division reproduces the draft.

Consumer advocates strongly supported the label.

The Division has verified that the five percentages sum to one hundred and certifies the label as the only document in the file that tells the shopper, in advance, what proportion of the page is the thing he asked for. The Division draws attention to the first line. On the proposed label’s own figures, organic results make up a little over a third of a search results page, which makes the answer to the shopper’s question a minority ingredient in the response to it — present in about the same proportion as the advertisements, and outnumbered by everything else combined.

The Division also commends the drafters for the fifth line, a category the food label lacks and has always needed.

Plaintiff’s Injury Reaches Recovery Stage

Three weeks after the original incident, reporters returned to the injured shopper. The Division reproduces the interview.

“How’s your thumb?”

The man raised it.

“Better.”

“What helped?”

“Mom kissed it.”

“Anything else?”

“I stopped scrolling.”

“Did you eventually find the product?”

“Yeah.”

“Did you buy it?”

“Yeah.”

“From Amazon?”

The man became quiet.

“…yeah.”

The plaintiffs’ attorney intervened.

“DO NOT ANSWER ANY MORE QUESTIONS.”

The Division regards the plaintiff’s final answer as the reason the case exists, and the reason it will be difficult to win. The cost of the sponsored placements was real. It was borne by him. It produced an injury, a second injury, a press conference, a class of approximately everybody, and an amicus brief. And at the end of it he bought the batteries from the platform that had made him scroll for them, because the batteries were there, and the delivery was fast, and the alternative was to start over somewhere else with a thumb that had already done enough. This is the defining property of a well-priced externality: it is set just below the point at which the person paying it will leave. The platform does not need the cost to be painless. It needs it to be a little cheaper than switching, and the Division notes that a thumb which has already scrolled thirty-one TSD to find a product is the last thumb on earth that wants to go and scroll thirty-one more somewhere else.

Settlement Negotiations Begin

Consumer advocates proposed several remedies. The Division reproduces the term sheet.

The injured customer proposed an additional requirement.

The consumer group rejected the proposal as humiliating. The customer did not withdraw it.

“I would use it.”

Amazon’s product team reportedly wrote something down. The plaintiffs’ attorney saw them do it.

“DON’T YOU FUCKING MONETIZE HIS MOTHER.”

The Division has reviewed the four terms on the plaintiffs’ sheet and notes that the fourth is introduced with the words or simply, and that the drafters were correct to introduce it that way. The first three terms regulate the advertising. The fourth regulates the distance, and is the only one of the four that would have prevented either boo-boo.

The Division files the attorney’s final objection as the most farsighted statement in the case. Mom Mode is, in the plaintiff’s conception, a remedy: a recognition, at the moment of injury, that an injury has occurred. In the product team’s notebook, the Division has no doubt, it is a placement. A screen that detects excessive scrolling and interrupts it with a prompt is an impression, and an impression delivered at a moment of peak emotional vulnerability, to a user whose thumb is already on the glass, is the most valuable impression an advertising business can sell. The attorney had been in the room for three weeks. He had learned how the platform thinks. He knew, before anyone at the table had said the word “sponsored,” what the notebook said.

Jury Hears Final Argument

Plaintiffs’ counsel approached the jury carrying a smartphone. The Division reproduces the closing argument from the courtroom recording, including the scrolls, which the court reporter transcribed at counsel’s request.

“This case isn’t about advertising.”

Scroll.

“It isn’t about commerce.”

Scroll.

“It isn’t even about Amazon.”

Scroll.

“This case is about the human thumb.”

Scroll.

“A remarkable biological instrument that helped our ancestors grip tools, build civilizations, create art, write language, and manipulate the physical world.”

Scroll.

“And what have we reduced it to?”

Scroll.

The attorney stopped. His thumb had developed a tiny red mark. He stared at it. The courtroom went silent.

“Counsel?”

The attorney’s lower lip trembled.

“YOUR HONOR, I GOT A BOO-BOO.”

“Oh, for fuck’s sake.”

The attorney reached for his phone.

“Who are you calling?”

He held the injured thumb carefully in the air.

“MY MOM.”

The Division records that the plaintiffs’ counsel had, by his own estimate, demonstrated the scroll in court more than four hundred times over the course of the trial, and that the injury therefore arrived on schedule. It also notes the closing argument’s central claim, which is correct. The opposable thumb is among the defining adaptations of the species. It made the tool, and the tool made nearly everything else. It has now been assigned, for several hours a day across most of the human population, the task of moving advertisements out of the way of the things it was trying to reach, and the Division can find no prior use of the instrument that was paid for by someone other than its owner.

Amazon corporate communications issued one final statement before close of business.

Twenty seconds later, the attorney received an Amazon notification.

The attorney stared at the screen. Then at the jury. Then at Amazon’s legal team.

Amazon’s lead counsel slowly put his head in his hands.

“OH, COME THE FUCK ON.”

The Division files the notification as the case’s final exhibit and the business model’s complete diagram, delivered without counsel’s involvement by the one party in the courtroom that had not been following the proceedings. The sequence it closes is short. An advertiser paid for position. The position was paid for, physically, by a thumb. The thumb was injured. The injury was detected — by the same systems that detect everything else a customer does — and the injury was sold, twenty seconds later, as an impression, to an advertiser of thumb braces. The externality did not leave the platform. It completed a loop through the customer’s body and came back as inventory. The Division notes that Amazon’s lead counsel understood this before anyone else in the room, and that his final statement, unlike the company’s three, was not reviewed by corporate communications, and is the only one of the four with which the Division is in complete agreement.

Closing Statement

At press time, the proposed class had been certified for 41 million named members, the Thumb Wellness initiative’s finger-alternation reminder had been quietly withdrawn after the plaintiffs moved to add index fingers to the class, and the lead plaintiff’s mother had received 2,300 messages from strangers asking her to look at their thumbs. Three competing retailers have announced that their search results contain fewer sponsored listings than Amazon’s. The Division has tested all three. Two are correct, and the third was found, on inspection, to have moved the advertisements into a carousel that scrolls horizontally, transferring the cost from the thumb’s vertical axis to its lateral one, which the ergonomics researchers have described as “technically a different injury.”