🎥 Video lessons in English are coming soon. They'll be added right here, on this same page — keep this link saved. Everything in them is already in the written lessons below.
This guide is the written version of the seven lessons of Sales Machine on Autopilot. The goal: sell digital products without ever showing your face, using short videos made with Video Studio (with Claude running the factory), a simple mobile page, and DM automation.
Currency. The original lessons were recorded with Brazilian numbers (R$). Amounts here are converted at roughly R$5.50 per US$1 and rounded. Real results are labeled "original" so you can see what actually happened.
Tools. The lessons use Brazilian platforms (Kiwify for checkout, Youze for DM automation). Wherever they appear, this guide names a global equivalent: Hotmart, Gumroad, Stan Store, Payhip or Lemon Squeezy for selling; ManyChat for Instagram DM automation; Cloudflare Pages for free page hosting. Companion products mentioned in the lessons (Digital Income Code, 20K Plan, Top 3 Niches) are optional deeper dives.
Disclaimer. Everything here describes one operation's real data and worked examples. It is not a promise of income. Results vary widely depending on niche, effort, consistency, offer and luck.
The core idea of the whole course: a viral video does not pay the bills by itself. The path that comes after the video does, and that path needs a lot of the right people walking through it. Volume plus funnel. One without the other does not add up.
Lesson 1: The Map: Where the Money Comes From and How the Funnel Works
The most expensive lesson
One of the instructor's Reels hit 1.4 million views and sold nothing of a product priced at about $54 (original: R$297). Why? Almost nobody who watched followed him, knew him, or had heard of the product. Cold viewers were asked to pay a premium price immediately. Views are not buyers: views are people walking past your store window. Most do not look, some look, fewer walk in, and very few buy.
He tried to fix it: he made 20 versions of the same video (only the first 3 seconds changed), created a cheap entry offer of about $9, then ran another test at about $12 on two different pages. Still zero sales, and the test page received only 5 visits. Those tests are opened in full in Lesson 5. The lesson: the path after the video matters more than the video, and that path has a name: a funnel.
The four sources of money when you do not show your face
Affiliate. You promote someone else's product and earn a commission per sale. Some products pay 70-80% to affiliates. You create nothing, support no one, build no page: you only bring people to the link. (Global platforms: Hotmart, ClickBank, Digistore24, ShareASale, Impact, or affiliate programs inside Gumroad/Payhip.)
Cheap own product. A guide, spreadsheet, step-by-step or short course at roughly $5 to $20. The margin is yours and you can build a product ladder on top. The instructor's own ladder: entry product, then order bump, then upsell.
Paid promotion. When a themed page grows, brands and creators pay to appear on it. He saw an anonymous page receive an offer of $4,000 for a single Reel, at just six months old. That is not typical for beginners, but it shows a themed page becomes an asset.
Platform money. YouTube pays for ads on your videos. In Brazil that pays little per thousand views, so selling a product matched to the video topic usually pays much more than waiting for ad revenue (Lesson 6). Rates are higher in English and Spanish markets.
A warning about a fifth thing: you will see people selling ready-made accounts (build a profile, grow it, sell it). It exists but violates Instagram's rules; you can lose the account or get scammed when being paid. This course does not teach it.
The math that changes your head
The official sales page got 60,237 visits in one week and sold roughly 500 units of the main product. That is under 1% conversion: about 8 sales per 1,000 visits. A page that sells every day converts below 1%. So the earlier test with 5 visits could not have produced a sale: the test did not lose, it never started. Your job is not to hope for virality; it is to get lots of the right people to your link.
You also do not need a big following. Examples seen while preparing the course: an anonymous profile made its first two sales with under 100 followers; another made its first sale at 82 followers; a profile with about 400 followers (all the right audience) made roughly $160 (original: R$900) in its first month. Four hundred right people beat 100,000 wrong ones. The number that matters is how many right-audience people reach your link.
The funnel running right now (8 steps)
Reel. Made in Video Studio with Claude, nothing recorded or hand-edited. It ends with one simple call to action: comment a keyword.
Comment. A DM automation tool (Youze in the lesson; ManyChat is the global equivalent) instantly sends the commenter a DM that asks them to follow you and delivers the link. Works while you sleep.
Guide page. A simple mobile page showing what the person will receive and the first step. It warms them up.
Official page. The sales video (the instructor's own recorded voice). The buy button appears only when the offer section starts.
Checkout. With an order bump: a checkbox to add another product for a small amount. In the sample week, 36% of buyers took one bump and 22% took the other.
Upsell. After paying, a bigger one-click offer (no card re-entry). This training is that upsell: 9% of buyers took it the same week.
None of the bump or upsell sales needed a new customer or another cent of ad spend: it is money from the same person in the same minute.
The factory and the clerk
The funnel needs volume; one video a week will not fill it. Video Studio is the factory: you ask Claude, it writes the script, picks the style, builds voice, images and captions, and the video comes out ready. Instead of one video a day you can make five, six, eight. Video Studio needs Claude to run: Claude thinks, the Studio is the factory floor. On the other side, the DM automation tool is the clerk who answers every commenter. You stay in the middle watching numbers.
The course roadmap
Lesson 2: choose what to sell, do the full profit math, build the profile.
Lesson 3: Reels that bring buyers, and how to spot a winner against your own average.
Lesson 4: capture, turning comments into DM contacts.
Lesson 5: conversion, with the instructor's real test results.
Lesson 6: YouTube without showing your face, and how to place links correctly.
Lesson 7: the numbers, finding leaks, scaling.
Do not skip lessons and do not hop strategies. People who change path every month get nowhere. People who stay on one path until it works, sell.
Action steps
Pick ONE of the four money sources. No product yet? Start with affiliate. Already have a product or idea? Start with a cheap own product.
Sketch your funnel on paper: video, comment, DM, page, checkout. Even if it is all empty.
Open Video Studio and ask Claude for your first video on any topic in the niche you are considering. You do not have to post it yet.
Lesson 2: What to Sell, What You Really Keep, and a Profile That Works as a Storefront
The opening trap: a 70% commission on a product costing about $17 (original: R$97). Does $11.90 land in your pocket? No, and that mistake makes people think they are earning while they are actually paying to work.
Pick a niche that sells, not just one that gets views
Documentaries, emotional stories and loose curiosities get views and sell little: people watch, feel something, and leave. What sells is pain. Four big areas: (1) money and income, (2) relationships, (3) health and weight loss, (4) mind, psychology and behavior. People with an annoying problem buy solutions.
Go narrower: the more specific the audience, the faster you sell. Anonymous profiles seen in research spoke only to women who want to leave a 9-to-5 and earn their own money, or only to mothers, or only to people recovering from a breakup. Small, precise audiences; they sold with a few hundred followers. Choose an audience similar to you or one you truly understand. If you are torn between two niches, the Top 3 Niches product was built for exactly that.
Option A: affiliate (fastest start)
Browse the marketplace of your affiliate platform (Kiwify in the lesson; Hotmart, ClickBank, Digistore24 and similar globally) filtered by your niche. Use five criteria:
High commission: 50% or more; ideally 70%+.
Low ticket: up to about $17. Cheap products need less thinking. A $180 product is a hard sell to someone who just met you in a Reel.
A good, ready sales page. Open it on your phone. Clear title? Proof? Explains what the buyer gets? A bad page wastes your traffic and it will look like your fault.
Proof it sells. Search the product name in the Meta Ad Library. If the producer has run ads for weeks or months, it is selling. Nobody keeps paying for ads on a product that does not sell.
More than one checkout. Many producers have the same product at several prices. For your first sale, use the cheapest link; a first sale removes enormous mental weight.
Do not promote fifteen products: two or three at most, all in your profile's topic (a weight-loss profile sells weight-loss products, never English courses). Products that can be shown (a spreadsheet, a guide, a tool, a step-by-step) sell better in video than a video course, which is hard to show in a 30-second Reel.
You can also be an affiliate of the instructor's own product: the Digital Income Code (English edition: Video Studio) is a validated product that made over 100,000 reais (about US$18,000) in 30 days. Page, sales video, checkout, order bump and upsell are already running. The commission is 80% with automatic approval. You post the same types of videos, the same comment keyword, the same DM automation, the same page, only with your own affiliate link. The affiliate link is in the lesson description.
The full math: Example 1, affiliate
Original numbers: a R$97 product, 70% commission, Kiwify fee 8.99% + R$2.49 per order. Translated to a global example: a $17 product, 70% commission, assuming a platform fee of about 9% + $0.50 per order (use your own platform's real fee).
Fee. $17 x 9% = $1.53, plus $0.50 = about $2.03. That leaves about $14.97.
Commission. It comes from what is left AFTER the fee, not from the full price. 70% of $14.97 is about $10.50, not $11.90. The affiliate marketplace already shows what you earn per sale: always use that number, never the sticker price.
Volume. 100 sales in a month = about $1,050 in commissions (you generated $1,700 in revenue for the producer, but your number is $1,050).
Refunds. In the instructor's operation, nearly 8% of sales of the Digital Income Code asked for a refund (39 refunds on 502 approved sales). Use 8%: of 100 sales, 8 come back and those commissions vanish. 92 x $10.50 is about $965.
Ads. If the 100 sales came organically from video, ad spend is zero. You only subtract tools (e.g., DM automation about $9 to $15/month, your Claude plan).
Example 2: cheap own product with bump and upsell
A guide for about $9 (original: R$47) with an order bump of about $5 (R$27):
Main product fee. $9 x 9% = $0.81 + $0.50 = $1.31. You keep about $7.69 per sale.
Bump. A bump rides in the same order, so it pays only the percentage fee, not the fixed fee again. A $5 bump leaves about $4.55.
Volume. 100 orders x $7.69 = $769. If 30 take the bump (conservative; the real week had 36% and 22%): 30 x $4.55 = about $137. Net total about $906 (gross about $1,050).
Refunds. Same 8%: about $833 remains.
Upsell. Add a $17 upsell that 9% of buyers take (the real number that week): about $124 more after refunds. Total roughly $955.
Side by side for 100 organic sales: affiliate about $965, own product with bump and upsell about $955. A near tie. These are example calculations; your numbers will differ but the logic does not. Affiliate is the fastest path to start organic.
Let Claude do the math in a minute. Give it the price, the commission, the platform fee and the refund rate, and ask for a profit table in several scenarios. Do this before choosing any product.
If you create your own product
It does not have to be big: a $5 guide or a $9 step-by-step. Tell Claude your niche and ask for 25 cheap product ideas that solve one pain for that audience. Claude builds the content with you, you upload it to your selling platform (Gumroad, Payhip, Stan Store, Hotmart, Lemon Squeezy, etc.). Warning: the product must truly solve the problem. If it does not, refunds climb past 8% and you saw what refunds do to profit.
The profile is your billboard
When a video reaches 100,000 people, some click your profile. It has two seconds to answer four questions: Where am I? Is this for me? Can I trust it? What do I do now?
Name (the bold name, not the @handle). Instagram uses it most for search, so put the phrase your audience would type (if the niche is side income, "side income" belongs there).
@handle. Simple, easy to remember and spell: a brand name plus topic, or a personal name plus topic.
Photo. You do not appear, so use a logo: clean, high contrast, readable when tiny. Claude can help you design the logo and name to match the look of your Studio videos.
Bio, three parts. Who it is for; what they gain by following; a call to action with an arrow pointing to the link. Example: "For people who want income without showing their face. Step-by-step videos every day. Start here (arrow down)". The bio talks about the person, not about you.
Link in bio goes to something of yours: the guide page, the official page, or your affiliate link. Never to another social network.
Three pinned posts: 1) start here, 2) proof (real result, real screenshot), 3) the offer.
Highlights: How it works, Proof, FAQs; and as an affiliate, free valuable content so people receive something before you ask for anything.
No proof yet? Use your own case: what you tested, what you learned, the screenshot of your first result, even if small. Never invent testimonials or create fake people with AI voices. Fake proof surfaces one day and takes the profile with it. Your first real sale becomes your first screenshot.
One niche only. Instagram reads your videos to decide what your account is about. If every video is about something different, it does not know whom to show you to.
A number to track from today: of every 100 people who visit your profile, how many follow? If 100 visit and nobody follows, the problem is not the video, it is the profile.
Action steps
Pick a pain-based niche.
Choose two or three products using the five criteria, or define your cheap own product.
Build the profit table with Claude (price, commission, fee, 8% refund).
Build the profile: name with keyword, logo, three-part bio, three pinned posts.
Lesson 3: Reels That Bring Buyers and How to Find Your Winners
One second of dead air in a 10-second video is 10% of the video, enough to kill it. And a video with a million views can bring fewer buyers than one with ten thousand.
1. Volume: every Reel is a lottery ticket
Instagram shows Reels mainly to people who do not follow you yet. Each Reel is a ticket with new people. Post one a week and you have one ticket; post six a day and you have 42. Creators post two, three, four a day; the instructor does five to eight and records and edits none. In Video Studio you tell Claude the topic, it writes the script, picks the style, builds voice, images and captions, and while one renders it already starts the next.
2. Direction: model what already worked
Find the outlier video: open 5 to 10 big profiles in your niche and check each one's average views. If a profile averages 100,000 views and one video has 1 million, that is 10x its own average: something worked (format, hook or structure). Note: compare with that profile's own average, not with the most-viewed video in the niche. Remember this; it is the heart of the lesson.
The format is replicated; the topic is occupied. Copy the way, not the video. If the outlier is a list of five mistakes with large on-screen text, make a list of five mistakes about your topic, in your own way. Same topic and same title and you compete against the owner, and the owner wins. Never download someone else's video and repost it as yours, not with credit and not with a changed song: it hurts reach, can get you reported, and you can lose the account. With the Studio you make a new one every time.
3. The hook: layered, in the first second
The first 3 seconds decide everything. Use a layered hook: three things at once from second one: big text on screen, the first spoken line, and a moving, eye-catching image, all saying the same thing. Nothing frozen at the start, no logo sliding in slowly.
If you are starting a profile, make your first nine videos in nine different formats on purpose (list, story, quick tip, before/after, question...). You do not know yet what your audience likes; the nine will tell you. Then double down on the winner.
4. Structures that sell
Quick tip: one problem, one solution, under 30 seconds. At the end: "Comment the word and I'll send you the rest."
List: "Three mistakes keeping you from making money." People stay to reach the last item.
Before and after: from point A to point B, then show the path.
Story of the person your product helps: start with their situation, they try, it falls apart, they think the problem is willpower, then discover the problem was something else, the tool appears, the result, and close with "comment the word". Tell it as a situation, not a testimonial. Never invent someone saying they lost 22 pounds in an AI voice: that is a fake testimonial and it is not taught here.
Curiosity question: "At what age did you find out you can do this?" A decor profile used this on a wood-care tip, passed 100,000 views and sold a lot, because the viewer feels they wasted time not knowing and comments.
5. Metrics: use your own ruler
You will hear "shares above X% means viral, saves above Y% will explode". Be careful: those thresholds change by niche, profile and even month. Comedy is shared far more than finance; recipes are saved more than motivation. Using someone else's ruler makes your good video look bad or your bad one look good. The right ruler is the average of your own profile:
Open your last 10 Reels.
For each, write down views, shares, saves, and followers gained from it. If you use comment-keyword automation, add a fifth: how many people commented the keyword.
Divide by views, expressed per 1,000 views (20,000 views and 120 shares = 6 shares per 1,000). Paste the numbers into Claude and ask for the table.
Average the ten. That is your ruler.
Compare each video to the ruler. Near average is normal and tells you nothing. Double or triple the average is a winner: repeat that one.
Example: your averages are 6 shares, 4 saves and 2 keyword comments per 1,000 views. A video appears with 18 shares per 1,000: three times your average. Even with fewer views than the rest, it is the winner, because viewers sent it to someone. The audience is telling you to make more of it.
The number that matters most for sales is the last one: keyword comments per 1,000 views, because that becomes DM contacts. A video with triple your average there is the one that will fill your funnel. Redo this every week with your 10 most recent videos; if the average rises, you are improving and the ruler rises with you.
Two more numbers, also compared with yourself: retention (how long people watch; if one video holds viewers to the end while your average holds them to halfway, study what it did differently at the start. That is why you cut all silence) and followers per 1,000 views (tells you whether the video brings the right audience; lots of views and no followers means curious people, not buyers). Likes barely matter and are not in the ruler.
Pre-post test: will the person stop? Comment? Save? Send to someone? If the answer is no to all, redo it. Claude can run this test with you: show the script and ask.
6. The call to action
ONE call per video. Never "like, comment, share and click the link" together: people do nothing when asked for four things. It changes by video type: for a tip, gift or tool, it is just "comment the word and I'll send it to your DM". For a story or curiosity, it can be "link in my bio or comment the word". Always say what they get: not "comment I WANT IT" but "comment GUIDE and I'll send you the step-by-step".
7. Trial Reels
When you post a Reel as a Trial Reel (Instagram's feature), it is not shown to your followers and does not enter your grid; it goes only to people who do not know you. Great for testing hooks without cluttering your profile.
The instructor did this with the 1.4M-view video: 20 versions, the whole video identical, only the first 3 seconds changed; 20 covers in a folder; posted as Trial Reels through the automation tool (up to 15 per day per profile in that tool, so 15 on day one and the rest scheduled for the next day), each with its own automation and its own code in the link. The honest part: those twenty versions sent almost nothing to the link. The page received 5 visits. Why? Trial Reels go to strangers who watch and maybe like but almost never stop to comment a keyword and click a link from someone they do not know. The view comes; the click does not.
The lesson: Trial Reels are for one thing: discovering which hook holds the most new people. Compare versions with the ruler. But sales come from people who already know you a little. When you find the winning hook in a test, post it on your normal profile, to followers too, with the call to action and automation. That is where clicks appear. This is "repeat the winner": you do not always need a new video, sometimes only the hook changes, and Video Studio with Claude makes twenty different hooks for the same video in minutes.
8. Final details
Music: a trending track, very low, under the voice. Voice loud, music almost gone. The video can then also be found through the music.
Safe zones: keep captions and text out of the right side and the bottom, where the app covers with like/comment buttons and the description. Important stuff stays in the middle. The Studio builds with this in mind.
Post caption: include the keyword your audience searches (for a side-income video, "side income" must be there). Hashtags barely help anymore. Keywords in the caption, on-screen text and profile name do.
Action steps
Find three outlier videos in your niche, compared with each profile's own average.
Ask Claude, in the Studio, for one video modeling each format with your topic.
Post at least two a day for one week.
At the end of the week, build the table of your last ten videos, find your average and mark the winner.
Lesson 4: Capture: Turning Every Comment Into a DM Contact
In the sample week, 92 sales came from organic Instagram (bio, Stories and DMs) with no ad spend, and 42 more came from YouTube. The main door to that path is one thing: a comment with a keyword.
Why not put the link in the caption or comments?
Links in Instagram captions and comments are not clickable. Where links work: the bio, Stories (link sticker) and DMs. Of those, the DM is the only one that reaches the person without them searching. A link in the bio is a door they must find on their own; a keyword comment is you going to them and handing over the link.
Why "comment the word" works: (1) the comment itself helps the video, since lots of comments signal that it holds people and Instagram shows it to more; (2) the person who comments raised their hand: "I want it", a real lead and not just a view; (3) each step removed means fewer drop-offs: no profile visit, no bio hunt, the link arrives.
The automation tool
The instructor used ManyChat, which works but became expensive at his volume (about $45/month with a contact limit; when the limit is exceeded the funnel stalls). He moved to a Brazilian tool, Youze, about $9/month with unlimited contacts and automations; in one night he moved 75 automations over. That is roughly $36 per month saved, about $430 a year. For an international audience, ManyChat is the standard Instagram comment-to-DM tool; check your plan's contact limit against your expected volume, or compare similar tools (e.g., Instachamp, Chatfuel, Inro) before committing.
The mistake that costs a day: he connected his Instagram to two automation tools at the same time. Result: nothing was sent; people commented and got nothing. Two automation tools on the same Instagram fight each other. Rule: ONE tool only. To switch, turn one off, remove its connection, refresh permissions, and only then connect the other.
The six rules of a good automation
Keep it simple. No ten-box flows with questions and branches; every extra step loses people. Comment the word, receive the message with the link. Done.
One automation per Reel. You know exactly which video generated each contact and you can send different links for different video types.
The keyword. Short, easy, related to the video (GUIDE, VIDEO, CODE). Register variations: with and without accents, upper and lower case, common typos, and the same word in another language. Good video travels the world, and foreign viewers comment and buy too. If a person mistypes a letter and gets nothing, you lost them.
Require follow. Before getting the link, the person must follow you. Your profile grows with the funnel: each comment becomes a contact and a follower, and followers see your next videos and Stories, where many sales close.
Do not reply publicly with the same sentence every time. When a video explodes you get a pile of identical replies that looks robotic. Keep the conversation in the DM.
The reminder. One hour later, the tool sends automatically: "So, did it work?" Many people open the link, get distracted and forget. The reminder brings them back.
What to write in the DM
Short: three lines at most. Thank them, say what they will receive, add a button. Example: "Here is the step-by-step I promised. Tap the button below to see it." The button text should make people want to click: "See the step-by-step", "I want the guide". No giant paragraphs; nobody reads walls of text in DMs. Ask Claude for five short message versions and five button texts, use one per automation, then see which gets the most clicks. If you sell more than one product, use one keyword per product so each person gets exactly what they asked for.
Which link goes in the message
If the video promised a gift, a tool or a step-by-step (something the person will receive), the link goes to the guide page, which delivers what was promised and shows the next step. For any other video (story, curiosity, news), the link goes straight to the official page with the sales video. At one point his automations were sending people to a direct sales page without warm-up: the person asked for a gift and landed on an offer, felt tricked and left. The video's promise and the link's page must match.
The tracking code (UTM)
At the end of the link, after the question mark, add a code (a UTM parameter) saying where the visitor came from. Put the video's keyword in it. When a sale lands in your checkout platform, you know it came from the GUIDE video or the VIDEO video. Without it you sell and do not know why. Claude builds these links for you, one per video, without typos. The code also reveals the quiet video that sells consistently: sometimes the video with the most comments is not the one that sells most, and a video with half the comments brings more purchases. Without the code you would repeat the wrong one.
Create an automation from scratch (7 steps, about a minute per Reel)
In your tool, create a new automation and select that one Reel only.
Type the keyword and all its variations (accents, caps, typos, other language).
Turn on require-follow.
Paste the short message and the button with the link: guide page if the video promised a gift, official page for everything else, with the keyword code on the end.
Turn on the one-hour reminder.
Activate it, and check that it is really ON (automations created in a paused state that nobody switches on are more common than you would think).
Test from another phone and account: comment the keyword and check that the message arrives with the right link. The instructor still does this today: one wrong letter in the link loses a whole day of contacts.
Bonus: some tools also publish and schedule posts (the tool in the lesson allows up to 15 per day per profile) so the video goes live with its keyword already configured.
An example of what happens in the DM (logic, not a promise)
One Reel gets 200 keyword comments. 100 people click the link right away (half). The one-hour reminder brings 20 more: 120 clicks. Twenty extra people from one message that sends itself, on every Reel, every day. You can see how many people received the message in the tool and how many visits came with that code on the page. If 200 received it and only 20 visited, the problem is the message: too long or the button unclear. Change it and look again.
Optional extra: ask for the email inside the DM before sending the link, so people who do not buy today stay on your list for later offers. Skip it at first; keep the idea for when the funnel is running.
Stories: where much of the selling closes
Reels bring new people; Stories close sales, because followers see them daily and already trust you a bit. Do not throw the link into every Story: it hurts reach and tires viewers. A five-part sequence works:
Interaction: a poll or question ("Do you want to make money without showing your face? Yes/No"). It makes Instagram show your Stories to more people.
A story: a problem or situation your audience lives.
Proof: a real result, a real screenshot (for example, a sale screenshot from your checkout platform). Nothing invented.
The objection: the doubt that blocks buying. Is it monthly? How do I access it? Do I need to show my face? Answer it before they ask.
The link, only at the end, with the link sticker and text: "Tap here to see how it works."
Claude writes this with you: tell it your product and audience, ask it to list that audience's pains, then ask for a Story sequence for each day of the month, one pain per day, with your product as the solution. A month of Stories planned in minutes. Ask the Studio for the Story images in the same look as your Reels.
Affiliates: everything works the same; only the link changes (your affiliate link instead of your page). Check whether you can add a tracking code to the end of it.
Three most common errors: (1) a hard keyword people misspell and get nothing, (2) a link that does not match the video's promise, (3) never testing the automation. All three are fixed in five minutes.
Action steps
Create your account on one automation tool (ManyChat is the global option) and connect Instagram with no other automation connected.
Pick the keyword of your next Reel and register all variations.
Build the automation with the seven steps above.
Test it from another phone.
Ask Claude for your Story sequence for the week.
Lesson 5: Conversion: Page, Sales Video, Checkout and Honest Test Results
Three offer tests, three times zero sales. That does not mean the offers were bad. It means something almost nobody tells you, and it can save you months of wrong decisions.
The page
People arrive from a DM on their phone, so build the page for mobile, not desktop. Keep it simple. Order on the official page: a clear headline saying what the person gets, who it is for, the sales video, what is inside the product, proof, and the button. That is all. No menu, no links to anywhere else, no distractions: every extra link is an exit.
You do not need to code. Claude builds the page: describe the product, audience and contents, and it assembles the whole page and can publish it. The instructor hosts for free on Cloudflare (Cloudflare Pages); Netlify and GitHub Pages work too. His folder holds several pages, one per test, all made this way. If you prefer a no-code route, Stan Store, Gumroad or Payhip product pages work as the page for a simple product.
The sales video
On the official page, the video is what sells most. The rule here: it is the instructor's own recorded voice, with the editing on top (examples running on screen, produced in Video Studio). Why a real voice? Because the person is about to trust you with money; a real voice builds trust. Also: the buy button appears only when the offer section arrives. Before that, the viewer watches, understands what it is, sees examples, and only then sees the price. If the button appears in second one, many click without understanding and abandon at checkout.
Exception: cheap products. At $5 to $7 you may not need a page; the link can go straight to checkout, because the checkout page already has the product image and what is included, so it works as a mini sales page. The cheaper the product, the less explanation it needs.
The checkout and the ladder
Order bump: a checkbox below the payment: "Add this for a small amount". The person already has the card in hand and has decided to buy; ticking one more box is easy. In the lesson, the Digital Income Code checkout offered the 20K Plan and the Top 3 Niches; in the sample week 36% ticked one and 22% the other.
One-click upsell: after paying, a page with a bigger offer that the buyer purchases in one click, no card re-entry. This training is that page. 9% took it. (Platforms with native one-click upsells: Hotmart, Kiwify; globally Stan Store, ThriveCart, SamCart, and Shopify apps. If yours lacks it, send buyers to a thank-you page with a discounted offer.)
A hole discovered the hard way: one customer paid, went to the upsell and the screen froze on "processing". How many others froze and never complained? One-click buttons sometimes hang because of bad internet or a slow bank, and the person gives up. His fix is a backup link: if the page sits on "processing" for 8 seconds, or errors, a text appears: "Button stuck? Tap here to finish your purchase." It leads to the normal checkout for the same product. Claude added it to every funnel page.
Proof on the page means real screenshots of sales, real results, real examples running. Never an invented testimonial. Right under the button, the guarantee in one sentence: if you do not like it, you can get your money back within the period.
Read the numbers of your page: seven numbers
In one week the official page got 60,237 visits. Reading the funnel, number by number:
Visitors: everyone who opened the page. About 60,000.
Started the video: about 34,000 (57%). And 21% reached 80% of the video.
Click: pressed the buy button. 6,515 (11% of visits).
Checkout: reached the payment screen. Every click leads there, but reaching checkout is not buying: in that week the platform recorded 1,319 abandoned carts just on the main product.
Purchase: about 500.
Conversion rate: purchases divided by the stage you are looking at. 500 purchases over 60,000 visits is under 1% (about 0.8%). 500 over 6,500 clicks is nearly 8%. Same funnel, both correct. Always say which stage you are dividing by.
Revenue and profit are different. Revenue is what came in: about 148,000 reais that week (roughly US$27,000), which is about $0.45 of revenue per visit. Profit is what remains after platform fees and refunds. Each sale of the main product (about $54) leaves about $49 after fees, and nearly 8% ask for a refund. High revenue with low profit is the most common trap.
Revenue per visit is the number that ties it together: divide what you sold by the visits. Here, every person who opens the page is worth about $0.45 on average. If you pay much less than that to bring each visit, there is room for fees, refunds and profit; if you pay more, there is none. And you can improve it without bringing one extra person, just by fixing the page.
The device leak. Split the numbers by device: people who opened the page from Instagram on Android started the video 75% of the time; from Instagram on iPhone only 32%. Nearly 33,000 iPhone visits, and most never saw the video. Since the video is what sells most, that was a huge leak, found only by splitting by device. So open your page on both iPhone and Android, from inside Instagram, and check the video appears, the play button is easy to find and it loads fast. Light page, fast-starting video, big button, easy to tap with a thumb.
The tests, with real numbers
Test 1. The 1.4M-view video did not sell the roughly-$54 product. His reading: cold audiences do not pay that much at once. So he created a roughly-$9 offer on the same product (not a new product, only a new price created in a few clicks), made the 20 video versions, posted them as Trial Reels, and built a page just for the test. Result: 5 visits, 1 click on the button, 0 sales.
Tests 2 and 3. A roughly-$12 price, split in two. In version A the buy button appears immediately; in version B it appears only near the end of the sales video. Two separate offers, each with its own link.
Version A: guide page 23 visits, official page 13 visits, 7 clicks on the button, 0 sales.
Version B: guide page 9 visits, official page 2 visits, 0 sales.
Which version won? Neither, because neither really played. The official page that sells daily converts about 8 sales per 1,000 visits. With 13 visits, you expect about a tenth of a sale: nothing. With 2 visits, even less. If one of the versions had sold once by luck, he would have thought it the winner, and been wrong.
Honest conclusion: these tests have no winner; traffic was insufficient. Two lessons:
Where the traffic came from. Trial Reels reach only non-followers. Trial Reel views barely become clicks: a million cold views can yield fewer clicks than ten thousand views from people who already know you.
Minimum volume before deciding. The practical rule:
Each version needs at least a few hundred page visits, ideally 1,000 per version.
Do the expected math: if your page converts near 1%, 1,000 visits give about 10 sales. If one version has 10 and the other 12, that is a tie. A difference of one or two sales is luck, not a result.
Let it run at least a full week, since some days sell more. In his operation the best hour was 11 AM and the worst was the middle of the night; a one-day test catches only a slice of the week.
Change one thing at a time. If you change price, page and video together and sales go up, you do not know what worked.
If you did not reach the volume, the answer is not to change the offer: it is to send more people (more videos with the keyword, more Stories with the link).
A trap that almost got him: when he made the "button at the end" version, the page he copied was already programmed to always show the button. Unless you remove that, version B equals A and you compare two identical things. After putting any page live, open it on your phone as a customer and check with your own eyes. Another smart habit: he never touched the original guide page. The test lives on a copy, so if it fails it does not break what already sells.
If you are an affiliate: simpler. The sales page and checkout belong to the producer; you only bring your link. Check two things: that the producer's page opens well on mobile from inside Instagram, and that you can add a tracking code to your link, so you know which video each sale came from.
Where the Studio fits: in everything that appears on the page: the examples in the sales video, demos, short clips showing the product working. Video Studio with Claude produces those, and it also produces the volume of Reels that delivers the 1,000 visits a test needs.
Action steps
Ask Claude for your page in the order shown (headline, who it is for, video, contents, proof, button).
Open it on iPhone and Android, from inside Instagram.
Set up your checkout with one order bump (and an upsell if your platform supports one), plus a backup checkout link.
Every week, write down the seven numbers: visitors, video starts, clicks, checkout, purchases, conversion, revenue and profit.
Lesson 6: YouTube Without Showing Your Face: Selling With Videos and Placing Links Correctly
One video on a faceless channel made over about $1,300 in sales (original: more than R$7,000) in a little over three weeks, more than double what the same video earned from YouTube ads. And there is a link mistake that makes many people lose these sales without noticing.
The math: why selling can beat ads
YouTube pays per thousand views, and in some markets (Brazil, in the lesson) that is low, far below the US, for example. You can have a video with hundreds of thousands of views and a small check at month-end. Instead, put in the video a product that fits the topic: an ebook, guide or step-by-step. One channel followed did exactly this: a psychology video about money, with a mid-video pitch for a cheap ebook, earning the sales above. Another channel with the same model made more than about $2,400 (original: R$13,000+) from a video with half a million views, and that channel was not even monetized.
Unmonetized channels still sell: YouTube does not pay you, but the sale lands in your checkout platform and you withdraw it. In the sample week, 42 sales came from YouTube with no ads behind them. (Note: in English and Spanish markets YouTube ads often pay considerably more per thousand views, so there it makes sense to add both streams: ad revenue and product sales.)
Warning: many of these channels grew by copying foreign videos: translating the title, copying the thumbnail and translating the script word for word. That took many of them down. YouTube calls it reused content and cuts monetization. This course does not teach that; it teaches how to model the right way.
The right way: the Viral Gap (3 steps)
Find the outlier video. Open a niche channel, go to the Popular videos tab, and compare views with that channel's own average and the video's age. If the channel averages 10,000 views and one video has 300,000, you found it. Often the channel did not even continue on that topic: an open door for you. Same ruler as Lesson 3: compare to the channel's own average, never to the biggest video on the platform.
Understand why the title worked. Viral titles often have five pieces: a high number, a concrete object you can picture, a secret (something hidden), a borrowed authority (someone who knows), and a useful fear (something to avoid). You do not copy the title; you see which pieces it uses and build yours. Example: "The 12 money habits a retired bank manager never tells his clients": 12 (high number), money habits (concrete thing), never tells (secret), retired bank manager (borrowed authority), and the useful fear hidden inside: losing money without knowing. Claude can generate ten versions with the pieces and you pick the strongest.
Change the angle, and if possible the country. That is the double gap: a topic that already proved it gets views, with an angle nobody has taken, in a language where nobody has done it yet. Example from the research: a topic translated literally into Portuguese got between 0 and 6 views. The same topic adapted to Italian, in local speech, passed 80,000. Adapted to German with a local authority, 27,000. Translating is lazy; adapting is what sells.
Before producing, a quick test: open YouTube in an incognito tab and search the title in the language you will post in. If a lot of channels already do it, skip. If nobody does, it is your turn.
Editing does not have to be cinema. Channels that sell well use simple editing: an image that matches what is being said, captions, a good voice. What sells is the script and the call to action. In Video Studio you pick the style and Claude assembles it with images that match each sentence.
The script: the pitch goes in the middle
The sales call does not go at the end of the video, where much of the audience has left. It goes in the middle, at the paragraph of greatest pain. Read your script and find the passage that hits the viewer's problem hardest. In a video about discipline, for example, the part about being unable to keep a routine when nobody is holding you accountable. There you place the bridge: a line like "this is why I need to tell you something", then you introduce the product that solves that exact pain (say, a discipline guide at about $6) and say where the link is.
The product must match the video topic: a discipline video sells a discipline guide, a relationship video sells a relationship product. And it must truly deliver what the pitch promises; otherwise refunds come, and you saw in Lesson 2 what refunds do to profit. Claude writes the bridge with you: show it the script and the product and ask it to find the highest-pain paragraph and write the pitch, then adjust it to your voice.
Where links actually work on YouTube
Long videos. Links are clickable in two places: the description and the pinned comment (the first comment, which you write yourself and pin to the top; the one most people see, especially on mobile). Use both. Pinned comment: short; a line about the pain, the product name and the link. Example: "If you cannot keep a routine when nobody is checking on you, the guide I mentioned in the video is here." Then the link. No ten links, no wall of text, one link with that video's code. In the description, put the same link on the very first line, before any other text, because that is what shows without the person expanding it.
Shorts are different. A link in a Short's description is NOT clickable, and a link in a Short's comments is NOT clickable. The person sees the address and cannot tap it. This has been true for a long time. So if you record a Short saying "link in the description", you are sending people nowhere. (Platform rules change, so check the current state before you rely on this.)
How does a Short sell? Two routes:
Related video. When you publish a Short, YouTube lets you pick a related video from your own channel. It appears as a little button inside the Short, and one tap takes the person to that video. Link every Short to a long video of yours on the same topic. The long video has the mid-script pitch, the description with the link and the pinned comment with the link. The Short is the bait; the long video is where the sale happens.
Channel links. In your channel profile you can add links that show in the header under the name. People who like a Short tap the channel name, land on your profile and see the link, like an Instagram bio. Put your guide page or official page there.
So a Short's call to action changes. No "link in the description". Say instead "the full video is below, tap the related video", or "the link is on the channel profile".
Always follow platform rules: a channel with no violation warnings, a link to an honest page that delivers what the video promises. No strange link shorteners, no misleading pages. If YouTube thinks a link is suspicious it may remove the comment or even the video.
The complete path
The Short, made in the Studio, pulls in new people.
The Short's related video leads to the long video on the same topic.
The long video has the pitch in the middle, at the highest-pain paragraph.
The viewer taps the link in the pinned comment or the description.
They land on the page, or straight on checkout if the product is cheap.
Whoever does not want the long video yet taps the channel name and sees the header link. Two paths, neither depends on a link that cannot be clicked.
Track every sale by video
As in Lesson 4, put a code at the end of the link. Long video 1 gets code 1, long video 2 gets code 2, and the channel header link gets its own code. In your checkout platform, filter sales by code and see how much each video sold. Then you know which script, topic and pitch sold, and you do more of that. Without it you have ten videos live, sell a little, and do not know which one pays the bills.
A simple per-video calculation: sales divided by views. Say a 20,000-view video sold 10 ebooks at about $6: about $60, roughly $3 per 1,000 views. Compare that with what YouTube ads pay per 1,000 views in your own analytics. That comparison shows why selling can pay more. Use your numbers, not mine.
Volume, other languages and the ladder
The same video you make for Reels can become a Short; the Studio already outputs vertical. Post to Instagram and YouTube. On Instagram the call is "comment the word" (DM automation sends the link). On YouTube the call is the related video or the channel link. Same factory, two storefronts.
Other languages. In English and Spanish, YouTube often pays more per thousand views, so adding ads plus product sales is worthwhile. The instructor's own Video Studio sales video exists in English and Spanish with the same examples, adapted, with pages live and dollar checkout for buyers abroad: the double gap applied to his own product.
Do not stop at the first product. A channel selling only one cheap ebook has a low ceiling. What makes the channel pay is the ladder: a cheap entry product, then a bigger product, then complementary ones. Someone who bought the first already trusts you, and it is much easier to sell a second product to a buyer than a first one to a stranger.
Who produces all this? Video Studio with Claude: long YouTube video with script, voice, images that match the narration, captions and the mid-video pitch. One long video a day becomes a possible routine, and with Shorts cut from the same script, each long video gets three or four baits pointing to it. To pick the channel's niche, the Top 3 Niches product helps; the 20K Plan helps build the product ladder.
Action steps
Find three outlier videos in your niche, compared with each channel's own average.
Test the title in an incognito tab.
Ask Claude for the script with the bridge at the highest-pain paragraph, and produce it in the Studio.
Put the coded link in the description (first line) and in the pinned comment.
Make three Shorts on the same topic, each with the related video pointing to the long video.
Add your link to the channel header.
Lesson 7: The Numbers: Finding Leaks and Scaling What Works
In one week, 207 people tried to pay by card and were declined. 207 people who wanted to buy. That does not show up anywhere unless you go looking. This final lesson puts everything together: the machine's numbers, where it leaks, and how to scale.
The six numbers
Every sales machine has the same sequence: views, comments, DMs, link clicks, people at checkout, sales. Each step to the next is a place where people leave. Look at these six numbers every week (the instructor does it every Friday) and find which transition leaks, because each leak has a different fix.
Five leaks
Lots of views, few comments. The video holds people but the call to action does not work: the keyword is hard, or people do not understand what they will get. Fix: a clearer single call that says what they gain ("Comment GUIDE and I'll send you the step-by-step"), and make sure it is not hidden under the app's buttons. Use the ruler from Lesson 3: compare keyword comments per 1,000 views with your own profile average. If it is below your average, the leak is here.
Good comments, few clicks. People comment, get the DM and do not click. The problem is the message: too long, button unclear, or they forgot. Fix: a short message, a button with clear text, and the one-hour reminder. Also check the automation is switched on.
Good clicks, few people seeing the offer. The person reaches the page and leaves. The page promises something different from the video, or is confusing, or slow on mobile. This was the biggest leak found on the instructor's page: Instagram on Android started the video 75% of the time; on iPhone only 32%. Over 32,000 iPhone visits and most never started the video. Fix: open the page on both devices from inside Instagram and make the watch button impossible to miss.
Full checkout, few sales. People arrive at checkout and do not pay. These are the 207 from the opening. In the sample week: 735 card attempts, 526 approved (72%), 207 declined (28%). On the local instant-payment method (Pix in Brazil): 403 generated, 293 paid (73%), 109 unpaid. Plus 1,319 abandoned carts on the main product. If half the declined cards had paid another way, that was over 100 more sales in the week. Fix: show an alternative payment method clearly as a second option for people whose card was declined (for an international audience: PayPal, Apple Pay/Google Pay, Link, Klarna/Afterpay, SEPA, iDEAL, local wallets depending on your market; choose a platform that supports them); use the backup link when the one-click button hangs; and mention the alternative methods in your Stories and on the page, because many people do not know they can pay that way.
Good sale, customer does not come back. They buy the first product and that is it. Fix: the ladder. Order bump at checkout, upsell after payment. In the sample week: 36% on one bump, 22% on the other, 9% on the upsell.
"It did not sell because it did not even have 100 clicks." The official page converts near 8 sales per 1,000 visits. At 20 visits you expect 0.1 sales: nothing. Many people abandon a product when they have not really tested it. Before changing product, niche or strategy, reach a few hundred visits, ideally 1,000. Only then does the number start to say something. That is exactly what happened with the roughly-$9 and roughly-$12 tests: 5 visits, 13 visits, 2 visits. None had a chance.
A worked example (logic, not a promise)
One Reel with 10,000 views. From it come 200 keyword comments. 150 people receive the DM and 120 click the link. 40 reach checkout. 4 buy a product of about $9. That is about $34 of revenue from a single video. Now improve the DM message and clicks go from 120 to 180: you sold more without a single extra view. And with an order bump, each of those sales is worth more. That is how you raise revenue by working on the numbers, not just hoping for a viral hit.
To have these numbers you need the code in the link, one per video. Without it you only know that you sold; with it you know what sold. In your checkout platform, filter sales by code: this video brought so many, that one none.
Two more things the week's numbers showed. Hour of day: the best selling hour was 11 AM (78 sales in that hour alone), the worst was 4 to 5 AM. So he knows when to post and when to send Stories. (Your best hour will depend on your audience's time zone: look at your own data.)
Scaling: three ways
Found the video that sells? Scale means multiplying what already works, not inventing new things.
Repeat the winner. Same video, only the hook changes. That is what he did with the 1.4M-view video: 20 versions, different first 3 seconds. Trial Reels help find the hook, but clicks arrive when you post the winner to people who already follow you too. Video Studio with Claude does this in minutes: you do not need a new video, you need a new hook.
More profiles in the same niche. If a niche sells on one profile, open another with a different name, face and visual line, selling the same type of product. Each profile is one more door into the same funnel. But only after the first one is selling; do not start with five idle profiles. Each new profile gets its own connection to the automation tool. Never two automation tools on one profile.
More languages. The double gap from Lesson 6: a topic that sells in your language, adapted (not translated) to English, Spanish, Italian. The same product opening another market.
The failures, and what each taught
The 1.4M-view video that did not sell the roughly-$54 product: views alone pay nothing, and cold audiences barely click.
The roughly-$9 and roughly-$12 tests: zero sales on 5, 13 and 2 visits. A test without volume has no winner; decide nothing before hundreds of visits per version.
Two automation tools connected at once: messages stopped going to everyone. One tool only.
The upsell frozen on "processing": always have a backup link.
Automations sending people to the wrong page: the video's promise and the link's page must match.
All these errors have one thing in common: he found them only because he was looking at the numbers. Whoever does not look repeats the mistake for months without knowing.
The routine of the machine
Every day: the Studio and Claude produce the videos and you post. Each video with its own automation and its own code in the link. Stories at the hour that sells most.
Every Friday: open your checkout platform and look at sales by code. Open your spreadsheet of the last ten videos and update the average. Look at the six funnel numbers. Find the leak. Change only one thing.
The following week: look again. Repeat the video that sold with a new hook. What did not sell, learn from and move on. Claude helps you read all of this: paste your numbers and ask where the leak is.
Where to go from here
The Digital Income Code is the base of all of this, the start of the path. The 20K Plan helps you reach the first big number. The Top 3 Niches helps you pick where to bet. The links are in the lesson description. Video Studio with Claude is the factory that makes the volume happen daily; without it you are back to hand-making videos one at a time and the machine stops.
If you want to start without creating a product, remember: you can become an affiliate of the instructor's own product, a validated one that made over 100,000 reais (about US$18,000) in 30 days. You copy the path he uses on his Instagram, step by step (videos, automation, page) and put your link in. Everything in these seven lessons works for promoting it. The affiliate link is in the lesson description.
The whole training in three questions
How does this make money? Videos that bring the right audience, a funnel that turns that audience into contacts, and a product that solves a real pain.
How does it become a sale? Through the path: Reel, comment, DM, page, checkout, order bump, upsell. Each step done well, each leak fixed.
How does revenue grow? Watching the numbers every week, recovering those who almost bought, repeating what sold, opening more profiles and more languages.
You do not need to show your face. You need method and consistency. Now you have the method. The consistency is up to you.
Action steps
Set up the six-number weekly review (views, comments, DMs, clicks, checkout, sales) in a simple spreadsheet; pick a fixed day for it.
Check your checkout for a visible alternative payment method and a backup link for any one-click upsell.
Do not change product, niche or strategy before each version has a few hundred visits (ideally 1,000).
Repeat your best video with three new hooks and post the winner to your normal profile.
Reminder: the numbers in this guide come from one operation's real week and from illustrative examples. They are not guarantees. Results vary and depend on your niche, effort, consistency, offer quality and market conditions.