Part of the guide: Network Marketing Lead Generation
I’ve been in this industry part-time since 2002. Over twenty years, I’ve watched hundreds of “opportunities” come through my inbox — supplements, skincare, travel clubs, coffee, crypto, prediction markets, AI tools, insurance overrides. Most of them were pitched to me the same way: comp plan first, product second, customer never.
So I stopped evaluating opportunities the way they wanted to be evaluated. I built one test instead, and I run everything through it. I call it the end-user customer test, and it’s one question with five parts:
Would a stranger who will never join this business buy this product, at this price, more than once?
That’s it. Everything else — the comp plan, the culture, the events, the founder’s story — is downstream of that answer.
In this guide
Why this one question does so much work
The FTC’s own business guidance on multi-level marketing keeps circling back to the same distinction: whether compensation is driven by real sales to people outside the recruiting structure, or by participants buying to qualify. Their guidance states plainly that a program where participants earn primarily from recruitment, rather than from sales of product to end users, has the hallmarks of a pyramid (FTC Business Guidance Concerning Multi-Level Marketing). Their consumer-facing material on MLMs and pyramid schemes tells shoppers to ask whether there’s genuine demand for the product from people who aren’t distributors.
Notice what that means. The regulator’s central question and the business question are the same question. That’s rare, and it’s useful. You don’t need to become a lawyer to protect yourself. You need to become a customer-demand detective.
Here’s the mechanism, because mechanisms beat statistics: in a structure with real outside customers, money enters from the edges — strangers pay for something they want. Distributor income is a slice of that inflow. In a structure without them, money enters from the middle — new participants pay in, and income is a redistribution of their money. The second structure can look identical to the first for a couple of years. It just can’t survive the moment recruiting slows, because there’s nothing underneath it.
The five parts of the test
Run these in order. Don’t skip to part five because the first four are boring.
1. Retail price sanity. Find the closest comparable product at a normal retailer. Not “comparable in the company’s language” — comparable in a stranger’s language. If the pitch is a $79 monthly greens powder, price the greens powders on a pharmacy shelf. You’re allowed to charge a premium for real differentiation. You are not allowed to invent differentiation to justify a margin that exists to fund the comp plan.
2. Reorder without qualification. Ask: what percentage of orders come from people who have no position, no autoship requirement, and no rank to maintain? If the company can’t or won’t tell you, that’s your answer. The mechanism to watch: when a product is genuinely good, reorders happen because it ran out. When it isn’t, reorders happen because a bonus was about to lapse.
3. The referral test. Would you send your own mother a link to the product page — not the opportunity page — with no explanation? If you need to wrap it in a story about freedom or a Zoom call before it makes sense, the product isn’t carrying the sale. You are.
4. The unwind test. Suppose recruiting stopped tomorrow — zero new distributors, forever. Does the existing customer base keep paying? Companies that survive this question are businesses. Companies that don’t are recruitment engines with a product-shaped receipt.
5. Where the disclosure lives. Legitimate offers make it easy to see the numbers. If a program is a business opportunity covered by the FTC’s Business Opportunity Rule, the seller has to give you a one-page disclosure document before you pay. MLMs are handled under separate guidance, but the principle transfers: ask for the income disclosure statement, and read the median, not the top line. If income claims are being made by promoters in your DMs, know that the FTC’s Endorsement Guides put real limits on typicality claims and undisclosed material connections.
You have been paid late before
You have watched the number on the dashboard not match the number in your account. What is different now is simple: it settles on a public contract you can check yourself, not a dashboard I control.
Read the contract →Most participants earn little to nothing. No income is guaranteed. You can verify every payout on the public smart contract.
Four worked examples
Example A: a $99/month supplement stack. Retail sanity: the closest shelf equivalent runs a third of that. Reorder without qualification: the company reports “customers,” but the definition includes distributors on autoship. Referral test: I can’t explain the price to my mother without mentioning the business. Unwind test: fails immediately. Verdict: recruitment engine. I’ve seen this exact shape a dozen times.
Example B: a $29/month software tool with a referral program. Retail sanity: competitors charge $25–$49 for similar functionality, so the price is market-set, not comp-plan-set. Reorder: most users never touch the referral program. Referral test: I’d send the product page cold. Unwind test: if recruiting stopped, the tool keeps billing because people use it Tuesday morning. Verdict: passes. This is why most of what’s on my recommended products page is a tool someone would buy without a comp plan attached.
Example C: a token, staking pool, or prediction-market “opportunity.” Ask the test honestly: who is the end user, and what did they consume? If the only way the position generates return is more positions being bought, there is no end user — there’s a queue. I wrote about how this pattern migrated out of supplements and into digital assets in the MLM playbook didn’t die, it bought crypto and a prediction market. The wrapper changed. The math didn’t.
Example D: a $180 skincare line with a real retail following. Retail sanity: comparable to department-store pricing, which is where its customers already shop. Reorder without qualification: strong, and the company can produce the split. Referral test: passes. Unwind test: mostly passes — revenue would dip but not collapse. Verdict: a real product inside an MLM structure. These exist. They’re just rarer than the pitch decks suggest, and the comp plan can still be a bad deal for you even when the product is good.
What to do when something fails the test
Don’t argue with the person who brought it to you. They usually believe it. Argument turns a math problem into a loyalty problem, and you’ll lose.
Instead, ask for the two artifacts: the income disclosure statement and the customer-versus-distributor sales split. Requesting documents is not an accusation. It’s what a professional does before signing anything — and it’s the same posture I describe in how to recruit professionals into network marketing, because professionals ask for paper.
If the answer is a deflection, you’ve learned everything you needed. Cross-reference what you’re seeing against the common MLM red flags and move on. And if you find yourself building a business that requires complexity to explain why the customer is real, read the complexity trap and the real customer — complexity is usually camouflage.
The uncomfortable part
This test will disqualify things you want to say yes to. It disqualified a lot for me, and it’s part of why I stepped back from doing network marketing seriously after 2017. Not because the industry can’t work — because I got tired of building on top of demand I couldn’t verify.
The upside is that once you have a test, you stop needing willpower. You don’t have to out-argue a good closer at 9pm on a Tuesday. You have five questions and a rule: if it fails part four, the answer is no, regardless of how much I like the people.
Build your business on strangers who buy because they want the thing. Everything else is a slower version of the same disappointment.
If you want the framework I use to build around real customers instead of around a comp plan, start here.
Three posts a day. One email a day.
Everything I publish, in one short daily email. Real numbers, no hype, unsubscribe in one click.
Get paid 100%, on-chain, the second the sale clears
Most affiliates earn little, and that is disclosed up front. The difference here is not a promise about what you will make. It is that you can verify every payout yourself, on a public contract, in about twelve seconds.
See how the payout works →Most participants earn little to nothing. No income is guaranteed. You can verify every payout on the public smart contract.
We ran a cruder version of this test in the nineties, we just called it “would a stranger buy it off a shelf.” Same five questions, fewer acronyms. What’s changed isn’t the test, it’s how good the wrapper has gotten at making a queue look like a customer base. Example C nails that part.
Sorry, Gus, but can I ask something probably obvious — when you say the wrapper got better at making a queue look like a customer base, do you mean literally that companies are getting sneakier on purpose, or is it more that everyone including the people running it genuinely believes their own numbers? Because that changes how worried I should be about missing it myself.
Ran the referral test on what I was selling back in year one. Failed it — I could not have sent my mother the product page without also explaining the compensation structure first, and once I noticed that I could not unnotice it. Took me until month nine to actually act on what that meant instead of just filing it away as an uncomfortable fact I’d deal with later. What finally moved me was watching a customer I actually liked go on autoship not because she loved the product but because dropping off would have cost her rank-adjacent people goodwill. That’s the unwind test failing in slow motion, not in theory.
Built a quick spreadsheet version of the 5-part test last night — dropdown for each of the 5 questions, weighted score, red flag if you fail part 4. Took maybe 40 minutes. Happy to describe how I set it up if anyone wants to build their own instead of running it in their head each time.
Trina, what weighting did you use for part 4? Feels categorical, not a gradient — averaging it in could hide a fatal result behind three passes.
Trina, this is a very useful thing you have built, thank you for sharing the method. I am wondering how part 1, the retail price sanity check, translates for a market like mine, because in Turkey the “closest comparable product at a normal retailer” is not always so easy to find for a category that is new here. Sometimes there is genuinely no shelf equivalent yet, and I am not sure if that makes the test harder to apply or if it is actually the answer itself.