Part of the guide: Affiliate Marketing
Most people build a quiz to get more leads. That is exactly why their quiz produces worse leads.
A quiz funnel is not a lead magnet with buttons. It is a sorting machine. Its job is to take a pile of strangers who all clicked the same ad and separate them into groups you will treat differently. If you are not treating them differently on the other side, you did not build a funnel. You built a longer opt-in form.
I have been in this industry part-time since 2002, and I spent 2007 making around 150 calls a day. Those calls were the quiz. Six or seven questions, asked out loud, and by question four I knew whether this person was a buyer, a builder, or a tire-kicker who wanted to talk about their cousin’s company. The software version should do the same work. Usually it does not, because of a handful of mistakes that all look like good marketing while you are making them.
In this guide
- Mistake 1: Your quiz flatters when it should filter
- Mistake 2: The segments do not change anything downstream
- Mistake 3: You score for identity when the money is in readiness
- Mistake 4: Your questions make an earnings claim on your behalf
- Mistake 5: The gate is in the wrong place
- Mistake 6: You harvest phone numbers you have not earned the right to use
- Mistake 7: The result page pitches instead of paying off
- Measure the number that costs you money
Mistake 1: Your quiz flatters when it should filter
“What kind of entrepreneur are you?” is a personality test wearing a funnel costume. Everybody finishes it. Everybody feels seen. Nobody gets sorted, because there is no wrong answer.
Compare that to a question with teeth: “How much can you invest in a business this quarter without asking anyone’s permission?” That question makes people uncomfortable, and roughly a third of them will bail. Good. The ones who answer it told you something a personality result never will.
Here is the thing. Completion rate is the metric that ruins quiz funnels, because the cheapest way to raise it is to stop asking anything real. A quiz that 90% of cold opportunity traffic completes is a quiz that learned nothing.
Mistake 2: The segments do not change anything downstream
This is the most common failure I see, and it is invisible from the inside. Someone builds four beautiful result archetypes. The Builder. The Freedom Seeker. The Side-Hustler. The Skeptic. Then all four get the same email sequence, the same webinar link, and the same call booking page.
If the segment does not change the next step, the segment is decoration.
Before you write a single question, write the destinations first. Three is plenty. For opportunity traffic, mine would be: someone who should buy a product and never hear the word “business,” someone who is ready to be recruited, and someone who is curious but not resourced yet. Those three people need genuinely different next steps. Sending the first one a business presentation is how you burn a customer who would have reordered for years, which is the whole argument in the complexity trap.
Build the three doors. Then build the questions that open them.
Mistake 3: You score for identity when the money is in readiness
Personality is fun to score. Readiness is what predicts behavior.
The four variables that actually moved my numbers across roughly ten thousand personal enrollments: available hours per week, available capital, decision authority, and prior experience selling anything to anyone. That last one is the sleeper. Someone who has closed a deal, run a territory, or managed a P&L will move faster than three enthusiastic beginners, which is the case I make in recruiting professionals.
Score those four. Weight them. Let the archetype be the wrapper on top, because people like being told what they are, but let the routing run on readiness underneath.
Mistake 4: Your questions make an earnings claim on your behalf
Watch this one closely, because it is the mistake that gets people in real trouble.
“How much do you want to be earning 90 days from now?” with options of $2,000, $5,000, and $10,000 is not a neutral question. It presupposes those outcomes are available. FTC staff guidance to multi-level marketers is direct about this: a company and its participants should not make claims, expressly or by implication, about earnings that are not typical for participants, and testimonials are generally read as representing results consumers can expect (FTC Business Guidance Concerning Multi-Level Marketing). A quiz answer choice is a claim you wrote and put in the prospect’s mouth.
The same applies to the results page and to any testimonial you drop under it. The FTC’s endorsement guidance covers what has to be disclosed about material connections and about results that are not typical (FTC’s Endorsement Guides: What People Are Asking). And if you are selling a business opportunity as such rather than an MLM position, there is a separate disclosure rule with a seven-day requirement attached to it (Selling a Work-at-Home or Other Business Opportunity?).
Ask about capacity instead of about fantasy. “How many hours a week can you protect for this?” tells you more than a dollar-figure daydream, and it does not put words in anyone’s mouth. If you want the longer version of what hype in this niche looks like, I catalogued it in MLM red flags.
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.
Mistake 5: The gate is in the wrong place
Two versions of this mistake, opposite directions, same cause.
Gate the email before question one and you lose the people who would have told you the truth. Gate nothing at all and you get insight with no way to follow up.
Put the gate after the last question and before the full result. By then the person has invested five answers, and the ask reads as delivery rather than toll. Show a partial result above the form so the payoff is visible and real. Something like “You scored in the top band for available time and the bottom band for available capital” is enough to prove you have something to say.
Mistake 6: You harvest phone numbers you have not earned the right to use
Opportunity traffic is where this gets ugly. People collect a phone field because a coach told them to, then hand the list to five uplines who dial it for a month.
If you email that list, CAN-SPAM applies to the commercial messages, including accurate headers, a working opt-out, and honoring it promptly (FTC CAN-SPAM Act compliance guide). If you call it, the Telemarketing Sales Rule and do-not-call obligations are live questions, and “they filled out a quiz” is not automatically the answer you think it is (Complying with the Telemarketing Sales Rule).
Practical version: make phone optional, label it for what you will do with it, and only ask the segment you actually intend to call. A phone number from your “not yet” pile is a liability, not an asset.
Mistake 7: The result page pitches instead of paying off
The moment after someone finishes a quiz is the highest-attention moment in the entire funnel, and most people spend it on a video sales letter.
Pay off the diagnosis first. Tell them what their answers mean, including the uncomfortable part. “Your capital score says do not spend money on ads yet” is a result that builds more trust than any bonus stack. Then give one next step, matched to their door.
One next step. Not a menu.
And build something real for the “not yet” pile, because on cold opportunity traffic that pile is the majority. A useful email course, a product they can actually use, a checklist. Those people either grow into buyers or they do not, but they will remember who told them the truth about their timing. That posture is most of why I stopped chasing volume and stepped back from network marketing as my main business after 2017.
Measure the number that costs you money
Quiz starts, completion rate, and opt-in percentage are all easy to move and none of them pay rent. Track one number per segment: what percentage of that segment took the next step you built for them within seven days.
Then kill the segments that never convert. A quiz with two doors that both work beats a quiz with six doors and one exit.
Open the tool you use, look at your result pages, and answer one question: does each result lead somewhere different? If the answer is no, you have a form, not a funnel. Fix that this week, before you spend another dollar on traffic.
If you want the ground floor version of how the pieces fit together, 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.
Mistake 3 is really an argument about predictive validity — personality typing has always had weaker predictive power for behavior than people assume, while stated capacity and prior experience are much closer to a criterion variable you can actually validate against an outcome. The article doesn’t use that language, but that’s the research distinction underneath “score for readiness, not identity,” and it’s worth knowing it has a name, because the same critique applies to most workplace personality assessments too.
The line about a third of people bailing on the uncomfortable capital question stopped me, because I’m fairly sure I’d be in that third. I’ve abandoned onboarding flows for less. Which makes me wonder whether this whole mistakes list is really a mistakes list for the builder, or a filtering test I’d personally fail on my way in.
Davey, notice you’re anchoring your own self-diagnosis on a number with no sample size or traffic source attached. “Roughly a third” isn’t a fixed property of the question — it needs a denominator before it means anything, including before you use it to decide which third you’d be in.
Raj, fair on the missing sample size, but I’ll say this — the diner used to hand out a comment card with one question that made about the same fraction of people put the pen down: “would you pay two dollars more for this.” Never had a number to back it either. Still told me more than the cards where everybody circled the top box.
We used to call this the money question and asked it out loud on the phone, not through a screen. Same effect — say a real number and a chunk of the room goes quiet. The quiz just moves the discomfort earlier and takes the human voice out of it, for better or worse.
In plain English, Gus: some questions make people uncomfortable enough to walk away, and that’s been true since long before anyone built a website to ask them.
Reading the four readiness variables against my own situation, one at a time, the way the article invites you to. Hours, yes, plenty of them now that I no longer have a classroom to prepare for each evening. Capital, modest but real, and I have been careful with what remains of it since the pension underperformed. Decision authority over my own choices, certainly — that much has not changed. Prior experience selling anything to anyone is the one that gave me genuine pause, because thirty years standing in front of a room and persuading children to care about a subject is not the same discipline as closing a deal with an adult who owes you nothing. I wonder how honestly a man my age is meant to answer that particular question, and what the number is meant to tell him once he has.
Arthur, I don’t think the honest answer to that question says anything about whether you’d do well at this — it only measures whether you’ve done this particular thing before. A quiz that scores you low there isn’t judging your character. It’s just admitting it can’t see the parts of you that thirty years in a classroom actually built.