The Screening That Screens Nobody

The Notice of Penalty Offenses Concerning Money-Making Opportunities went to more than 1,100 businesses on 26 October 2021, under 15 U.S.C. § 45(m)(1)(A). It enumerates eight practices the Federal Trade Commission has already held unlawful in its own decisions. Three of them describe the first thing a buyer is told about themselves.

• Offence two: "falsely representing that sales are limited to a limited number of buyers."
• Offence three: "falsely representing that prospects will be screened or evaluated for suitability."
• Offence five: "falsely representing that a prospect must act immediately."

Those three are one move in three tenses. The intake is capped, the prospect has been measured against the cap, and the measuring closes soon.

The notice carries no dollar figure and no date on its face. The date above is the Commission's own, from the press release announcing the notice on 26 October 2021, and it belongs to this notice rather than to the separate notice about fake reviews and misleading endorsements that went to more than 700 companies on 13 October 2021. The two are routinely merged, because they carry the same penalty figure between them.

What the notice does is arm a number that appears nowhere inside it. A business that has received it and then engages in one of the enumerated practices is exposed to civil penalties of up to $53,088 per violation under 16 C.F.R. § 1.98, for penalties assessed after 17 January 2025, and the Commission has adjusted that figure each January since 2016. It is a ceiling rather than a collection. What was actually collected in the matters this volume assembles is Chapter 4's table, and it is printed there in two columns for the reason that a ceiling and a recovery are different numbers.

Now the word that governs all three offences, and the problem it hands the reader. Each one turns on falsely. A false-representation offence is proved by setting a claim beside a fact. The claim that a prospect will be screened for suitability is a claim about a process, and two facts would settle it: the criterion the process applies, and the number of prospects it turned away. As of September 2026 this desk located neither of them, from any seller of a course, a coaching product or a done-for-you business opportunity: no published rejection criterion, and no decline rate.

The same exercise can be run on the other two, and it produces a short list of documents that nobody has ever been asked for. Each row below takes one offence from the notice of 26 October 2021, names the move in the sequence it describes, and names what a seller would have to hand over for the representation to be checked rather than accepted.

Every entry in the right-hand column is a thing a seller already holds. A business that caps an intake knows the cap. A business that declines applicants knows how many. A business that expires an offer knows the date it set. None of the six is a trade secret, a projection or an estimate, and none of the six was published by any seller in this record.

So the practice each offence describes is one that cannot be checked from outside by anybody who is not the seller. That is a reason to read the sequence instead of the claim. The sequence can be read, because it was pleaded in a federal court with numbered paragraphs, and what it shows is where the claim of selection sits in the order of events and what else is happening in the same minute.

The notice issued 26 October 2021 also decided something by listing screening at all. It sets the claim of a filter beside the claim of earnings, under the same statute and carrying the same penalty exposure. On the Commission's reading, telling a prospect they were evaluated is not sales atmosphere sitting outside the reach of proof. It is a representation of fact, and it either has a document behind it or it does not.

Where the application should be

The artifact this market opens with is a form headed Application. This book cannot tell you what any particular one of them asks. No seller in the record assembled here published the form, the criteria behind it or the outcome of any submission to it, and describing a document nobody has produced would be an invention of the kind this volume exists to price.

What the federal record does contain is one matter's version, pleaded under a docket number. The First Amended Complaint in FTC v. Growth Cave, LLC, No. 2:25-cv-01115-DOC(RAOx) (C.D. Cal.), filed 9 May 2025, describes an opt-in page that takes a name and an email address, near-daily email afterwards, and a video the prospect is instructed to watch before a call. Nothing in that stage puts a question to the prospect whose answer could disqualify them. A name and an email address are not facts about suitability. They are the delivery details.

The sales video did put criteria in front of the prospect, and they are on the record. The same complaint filed 9 May 2025 quotes the list from the KBA sales video describing who the program is for: "You want help launching your own online course to a consistent $10K-50K+ per month as quickly as possible," "You want to launch with as little risk as possible," and "You don't have the time or interest in stumbling through months of complex, expensive guesswork trying to do this on your own."

Read that list as a screen and check what it filters out. Every item is a statement about what the prospect wants. Not one is a statement about what the prospect has — no capital floor, no hours per week, no prior experience, no equipment, no test, no threshold of any kind. A criterion that a reader satisfies by the act of having read the advertisement is not a criterion. It is a mirror. Whatever qualification is happening there is being performed by the prospect, on themselves, in the seller's words.

That is a reading of one document rather than a finding about a market. The finding about the market is the absence around it: no seller in this record published a criterion that could exclude anybody.

The conversation that is both procedures at once

That is the seller's own account of what the strategy call is for, as pleaded in the First Amended Complaint filed 9 May 2025. Offence three of the notice issued 26 October 2021 covers falsely representing that prospects will be screened or evaluated for suitability, and the answer to the pitch is not that no evaluation occurred. Nobody outside the call can say. The answer is what else the same call was doing while it occurred.

Paragraph 46 of that complaint records that the price of $3,500 to $9,800 was disclosed to the buyer for the first time on that call, and that payment, or an application for third-party financing, was taken during the same call. The conversation described as an evaluation of the buyer is the conversation in which the buyer is told the price and pays it.

Set those two facts side by side and the difficulty is structural rather than moral. One person is running both procedures in one call, and the two procedures produce the same artifact: a payment, or no payment. A prospect who was screened out and a prospect who declined to buy leave behind identical evidence, which is none. No document is generated by the first outcome that is not also generated by the second, and no third party is present to tell them apart.

A screen produces rejections, and rejections leave records: a declined application, a returned deposit, a note on a file, a number in a monthly total. This desk located no such record and no such number anywhere in this market. A screen with no published criterion and no published rejection rate is not a screen that has been tested and passed. It is a claim in the same position as the earnings claim — an assertion whose supporting document the seller has never been asked to produce.

The video sits one step earlier and is worth naming for what it is rather than what it contains. The complaint filed 9 May 2025 records that the prospect is instructed, in capitals, to watch it before the call. Watching it is the only requirement placed on anybody before the money moves, and the pass mark is attendance. What that stage measures is compliance. It establishes that the prospect can be told to do something and will do it, which is a real finding about a prospect and is not a finding about suitability for a business.

Count what the sequence in that filing actually collects from the buyer before the payment, and the tally is short. It collects an email address, a viewing, and an appearance on a call at a time the seller set. All three are acts of compliance, and all three are free to supply. It collects no fact about capability at all, and no answer that a person could get wrong. A process that gathers three compliance signals and no capability facts has been built to measure something, and what it measures is whether the prospect keeps following instructions up to the point where an instruction costs money.

The gate that does exist

There is one gate in this transaction with a published pass mark. The complaint filed 9 May 2025 describes purchasers who had to score 100% on every quiz in the program in order to progress, and who were ordinarily not told the correct answers.

Note where that gate sits in the order. It is after the payment. A buyer who fails it has already paid between $3,500 and $9,800, per paragraph 46 of the complaint filed 9 May 2025, and failing it does not return the money. A test that must be answered perfectly, with no answer key supplied, has two outputs. The first is a customer who does not progress. The second is a record showing that the customer did not complete the program, and the second output is a document the seller can produce afterwards, to a complaining buyer, to a payment processor, or to anybody else who asks what went wrong.

So the transaction contains a screen after all. It runs after the money has moved, it is pointed at customers rather than at applicants, and its manufactured failure condition doubles as the seller's defence. Every element of it is verifiable from the pleading. None of it is the thing the word screening was used to mean at the top of the funnel.

The arithmetic of scarcity

Offence two of the notice issued 26 October 2021 covers falsely representing that sales are limited to a limited number of buyers. Checking it requires one number, which is the size of the roster, and sellers of this kind do not publish rosters.

One roster has become visible from outside, and it did so through a breach rather than through a disclosure. Coverage by PCMag, Bitdefender, Hackread and Indy100 of a breach of the online-course operation "The Real World," run by Andrew Tate, which ran between 21 and 26 November 2024, put the exposure at between 800,000 and 1,000,000 user records together with chat logs.

Four caveats travel with that range and none of them is small. The operation is not a seller of a YouTube course, and nothing in the federal record assembled for this book connects it to any matter in this volume. The figure counts user records, which is not the same as counting people who were paying on the day of the breach. The range exists because the outlets reporting it did not converge on one number. And a breach is not an audit; nobody chose the methodology, and nobody can be asked what was excluded.

What survives all four is narrow, and it is the only observation of its kind. On the one occasion when a roster in this industry was counted by somebody other than the seller, the count ran into the hundreds of thousands. That does not disprove any particular claim that an intake is capped, because no seller in this record published a cohort size and therefore no cap in this record can be tested against anything. It is the single external data point in a market that otherwise supplies none, and it is the number to hold in mind while a cap is being described to you.

A second reading of the same range belongs beside the first. A roster in the hundreds of thousands is not itself misconduct, and a large customer base is what a product that sells looks like. What it sits badly with is a claim about the intake. A seller who says the intake is capped while holding a roster of that order is describing a subset, and the subset is defined by the seller, counted by the seller and disclosed to nobody. That is not a cap a buyer can fail to make. It is a cap whose only evidence outside the seller's own records is the sentence that announced it.

Offence five, the deadline, is the cheapest of the three to test and the reader can do it alone. Write down the date the offer expires. Wait a week past it and ask for the same offer. The result is not evidence about the seller's honesty in any general sense, and this book does not claim it is. It is one dated observation, made by you, filed in a market where no observations of this kind are published by anybody.

The room the survivors are in

The last stage is a room, and the record describes what happens inside one of them. Paragraph 73 of the complaint filed 9 May 2025 states that when buyers complained on the seller's internal message boards, the Growth Cave defendants routinely deleted the complaints and any other negative posts.

Read that as a measurement problem rather than as a discourtesy. The best evidence available to the next buyer about what the product does is the visible experience of the people already inside it. Remove the negative posts and the people who wrote them go quiet or leave. The visible population then converges on the people for whom it went well, and it converges there without anybody fabricating a single testimonial. Survivorship of that kind is not a side effect of a community. It is the output.

Skool's pricing page, retrieved in September 2026 at United States pricing, carries a built-in affiliate program on both of its tiers, and that line is the one to read twice. On that page as it stood in September 2026, a member of the room can be paid for bringing in the next member of the room. Nothing about that makes any particular community dishonest, and no seller named in this volume is alleged to have run one on that platform — the deleted posts at paragraph 73 of the complaint filed 9 May 2025 were on the seller's own message boards, not on any named product. What it means is narrower and it is enough: the visible enthusiasm of an existing member is not, by itself, evidence about the thing being sold, because the structure permits that enthusiasm to be compensated and does not require it to be labelled to the next buyer.

There is a rule pointed at exactly that, and it postdates most of the conduct in this book. Section 465.5 of the Rule on the Use of Consumer Reviews and Testimonials, effective 21 October 2024, reaches insider reviews and testimonials given without clear and conspicuous disclosure of the connection. The operative word is the connection, not the relationship's name. An affiliate who is paid on referrals is connected, and so is an employee, and so is a lender to the seller's customers. The question that follows from the rule is one sentence long and is asked about a specific person in a specific video, by name.

The screen that does have published criteria

There is exactly one filter in this transaction whose criteria a reader can hold in their hand before any money moves, and it belongs to the buyer. Appendix B is that filter, printed in full so that it can be run rather than admired: what you are buying written down as an output rather than an outcome, a ceiling written down with the date you set it, an acceptance test written in observable terms before the work starts, a search of the seller's names against the Commission's own case listings, and a dated screenshot of the sales page and the guarantee.

Two of those items exist because of what this chapter has taken apart. The ceiling is written down first because the number will otherwise be settled inside the conversation described at paragraph 46 of the complaint filed 9 May 2025, in which the price is disclosed and the payment is taken in the same call. The search of the person as well as the company is there because, across the matters this volume assembles between 16 November 2022 and 6 July 2026, rebranding after an action is the ordinary course rather than the exception.

That reversal is the whole of the chapter's method. The seller's filter has no published criterion and produces no visible rejection. The buyer's filter has both, costs fifteen minutes, and is the only one of the two that anybody can audit.

What this chapter cannot count

This chapter counts nobody, and the reason is two missing denominators rather than a decision to leave the buyers out.

The first is the number of applicants any seller turned away, which is the subject of the chapter and which nobody has published. The second is the number of buyers who got what they paid for, which nobody has published either — not a seller, not a platform, not an academic, not a trade body. Every measured buyer outcome anywhere in this book was produced by a regulator counting a cohort while it was suing the seller, and those counts are printed in Chapter 4 because that is where the litigation is.

A chapter about a screen therefore has documents to read and no rejects to interview. Nobody was ever recorded being rejected.

What to send, and what to keep

Ask the seller, in writing, before any payment: how many applications did you decline last month, and on what criterion? It is question eight of the nine in Appendix B, and it goes in the same message as the other eight, because a question with a rule beside it gets a different answer from a question without one.

Then keep what comes back, with the date on it.

A number with a criterion attached can be set against the seller's own advertising and against the two facts that would prove offence three false, which is more evidence than exists in public today. A refusal to answer in writing is a finding of a different kind: it says the number exists somewhere in the seller's records and is not one the seller will put in an email. And silence is the third answer. If there is no reply, that is the reply, and it is a dated one as long as you file the message you sent.

The instrument is homemade, and the admission belongs here rather than in a footnote. No seller in this record has published a decline rate. No public dataset shows how many applicants any of these programs turned away. The four figures a buyer would need in order to test a claim of selection — applications received, applications declined, the criterion applied, and the date that criterion last changed — appear nowhere this desk could reach as of September

1. What would close the gap is one seller publishing those four figures for one named

month, or one regulator obtaining them in discovery and putting them in a filing. Until either happens, your own written question and the dated reply, or the dated silence, are the only record of the screen that anybody holds.