What all of them do that none of them teach

How this was built

Each dossier in this series is researched against the same brief, and that brief carries a standing instruction: after every three to five subjects, run a cross-dossier pass asking what do all of them do that none of them teach? Every subject below has a written cross-reference block in which the researcher recorded, pattern by pattern, whether that operation confirms or contradicts it, citing evidence from their own file.

Sample size matters here. A pattern that looks like a clean split between network marketing and affiliate files at ten subjects can vanish at 73, which is why every tally on this page is computed from the full grid rather than typed.

Nothing here is inferred from a dossier that never reached a verdict. Where a researcher's evidence did not bear on a pattern they wrote NOT ASSESSED, and those cells are marked · below rather than filled in with a guess.

The verdict grid

73 subjects, seven patterns, as recorded at research time.

✓ confirms  ✕ contradicts or inverts  ~ mixed or partial  · not applicable  · MLM network marketing   AFF affiliate/creator   PRG program operators

1. Authored the asset their field distributes. The thing the downline, audience or buyer passes around was written by the subject, not licensed from a company.
2. Owns story, tool, funnel, traffic and meaning — never just an audience.
3. Disclosure beats concealment. The operators with the cleanest disclosure records are not the smallest earners.
4. Platform independence at scale, branded-search dominance when small.
5. Corporate funnel quality is half the field result. The subject controls the click; somebody else controls everything after it.
6. The coaching/product layer is the life raft when the primary income source moves.
7. The field is exit liquidity. The people trying to become you are the market you eventually sell to.

The one that holds

Pattern 1 confirms on 60 of 73 subjects, in cohorts that share almost nothing else — a network-marketing leader in India, a travel blogger who refuses to sell anything, an AI-tools curator, a Filipino education channel, a fraud investigator funded by rental property.

What varies is what the asset is. For Lo Gatto it is a book and a 110-lesson course. For Schlappig it is 21,576 posts and roughly 43.9 million words. For Wolfe it is an index of 4,233 hand-reviewed tools. For Wardrop it is a CRM snapshot his signups import and run. In every case the subject wrote the thing their field hands to other people.

It is the closest thing in the series to a transferable instruction, and it is also the one almost nobody teaches, because it is slow and cannot be bought.

Pattern 5 is the runner-up, confirming 44 times. Whoever controls the page after the click controls roughly half the result, and most operators in this series chose their offer without ever auditing it.

Where the framework breaks

Five of the seven fail more often than they hold, and two of those failures matter more than the rest.

The product layer is not reliably a life raft

Pattern 6 says build the coaching or product layer, because when the algorithm or the comp plan moves it is what keeps you afloat. It is the most commonly taught idea in this industry.

Across 73 subjects it fails about as often as it holds. And it fails often in both of the cohorts people assume are opposites:

The failures are not near-misses, and they take four distinct shapes:

• Never built one. Aron Parker sells free trainings and routes everything to affiliate links. When his program collapsed there was no raft, and his own file records "we were not in a good spot financially." Ada Caballero runs no course, no book, no list.
• Built it welded to the vehicle. David Wood's product layer was Empower Network, so it died with the company and a three-year gap followed in which no venture is locatable in any source.
• Built it, then dismantled it on purpose. Jon Dykstra retired a $497 flagship and de-listed his own products mid-collapse. Matt Wolfe killed a working paid membership. Jason Wardrop liquidated his ladder into free affiliate bonuses. Pat Flynn had the most developed product layer imaginable — 17+ courses, a $99/month membership, four books, a 945-episode podcast — and sold it.
• Built it and it was not enough. Tracey Walker has a $9/month community with 131 members producing roughly $1,002 a month net. It exists, it works, and it is not keeping her afloat.

What actually separates the survivors is not their cohort. It is whether the product layer can outlive the vehicle that fed it. Worre, Higdon, Brookes, Ceballos and Dillard each built something that kept earning after the field income moved. Wood's could not exist without Empower. Parker's never existed. Justin Prince had a book and four courses and still reported no income during litigation while spending, in his words, "$100 grand a month" on lawyers — what rescued him was founder equity, not the product layer.

Hayley Hobson runs it backwards, and hers may be the most useful file in the series on this point: the mature residual is the raft, and she spent it buying a new audience. A settled income stream is the cheapest venture capital available, and most operators spend it defending a downline instead.

The instruction, restated honestly: a product layer is worth building only if it would still earn after the thing that currently pays you stops. Most of the ones in this series would not have, and the people teaching you to build one rarely make that distinction.

Platform independence is mostly aspirational

Pattern 4 is contradicted more often than any other pattern: 33 times, against 23 confirmations. And several files contain the stress test rather than the theory: Pera University lost four years of library to strikes, Mikael Møller's channel was removed for eight days, Adam Enfroy deleted 464 of his own videos. Graham Stephan has 5.18 million subscribers and a 41,000-person email list — a 0.5% capture rate.

A recurring signature across these files is a loyal audience attached to dead distribution: like-rates holding or rising while views collapse. The operators who survived it are the ones who captured contacts on the way up.

Disclosure is claimed more than it is practiced

Pattern 3 holds cleanly in only a handful of files and is mixed in more than any other pattern. The recurring shape is a written policy better than its execution — Matt Wolfe's file puts it plainly, that he wrote a better disclosure policy than he follows. The exception is instructive: Graham Stephan's disclosure record was clean enough that he settled an enforcement action for a reported $10,000 where peers in the same tranche paid $122,000 and $180,000. Cheaper exposure is the measurable dividend.

What the contradictions produced instead

The cross-reference blocks also surface roughly two dozen named mechanics that no pattern in the seven describes. They are candidates, each observed in one or two files, and they are listed rather than promoted because one observation is not a finding. The most testable:

• The funded proposal (Dillard) — price the front end so its revenue pays for the traffic that found the buyer. The purchase is the qualification event.
• The inevitable purchase (Basit Ali) — take commission on the one thing the audience must buy to act on any advice, then bundle your own product beneath it so the affiliate path is strictly cheaper. The product becomes bait rather than raft.
• Own the room (Worre, and in a different form Flynn) — physical venue or event ownership as an appreciating asset. Flynn's version charges the supply side for proximity to an audience, which no algorithm or comp-plan revision can take away.
• Honesty as a conversion mechanism (Møller) — when the monetized event is a free signup rather than a purchase, deflating expectations lowers churn and raises earnings per click. The testable generalization: operators with the worst claim hygiene may simply be the ones whose money depends on a sale.
• The platform-constraint product (Chrisp) — sell the content the distribution platform structurally refuses to distribute, so free and paid cannot cannibalise each other.
• Governance as a moat (Ceballos) — a written no-compete constitution inside an owned community. Free to install, compounds with size.
• The credential is a depreciating asset (Walker) — a company-issued distinction from 2014 now re-anchors her to the category she renounced.
• Language-pair arbitrage (Tazkeer, Pera University) — the opportunity is not the topic, it is the topic-language pair. Noted twice independently. In both cases the channel's titles point to one topic and its transcripts to another.

One further candidate — that notoriety is a cheaper input than trust — still has only three observations behind it and stays out of the grid on the same principle.

What to do with this

Four things follow, in descending order of how confident the evidence lets us be:

1. Write the thing your field will distribute. The only instruction here that nearly every documented subject supports.
2. Audit the offer's own funnel before you promote it. Pattern 5 confirms 44 times: whoever owns the page after the click owns about half your result.
3. Build a product layer only if it would outlive the thing that currently pays you. Not because someone told you to build a course. The distinction is the whole finding.
4. Capture contacts before you need them. Platform independence fails more often than it holds, and the files that survived a platform event are the ones that had somewhere else to send people.

Every claim above traces to a named dossier, and every contradiction is recorded in that subject's own file. The contradictions are the reason this page is worth reading, and the reason it is rechecked as the series grows.

← Browse the dossiers · How we verify · Report an error