## 1. Executive Summary Diogo Arnold lays out how he builds offers that survive cold traffic: research the market before writing a word, model the ads already scaling instead of inventing, sell the offer before the product exists, and wrap everything in a single mechanism delivered by a short, ugly VSL that exists mainly to buy data. His doctrine is that copy is the 80/20 of the business and traffic is only the amplifier one hit VSL can move R$100,000 a day so the work happens in the research and the message, not in ad hacks. The episode walks the full sequence: how he mines competitor swipe and transcript searches for what is already converting, how he maps the prospect's existing beliefs so the mechanism flies under them instead of fighting them, the exact structure of his 1,500 2,500-word mechanism, why he tests creative angles before writing the lead at all, and how a deliberately unpolished PowerPoint VSL let him validate an offer spending R$2,000 to sell R$7,000. ## 2. Alpha / Sauce - what actually moves the needle, and exactly how to run it - **Model what is already scaling; never innovate first.** Pull the five most-scaled cold-traffic ads in your market and copy their pains, promises, and mechanisms people have already paid to validate those structures. He traces this to Claude Hopkins' 1904 "scientific advertising": headlines and structure have been a science for over a century, and combining a proven headline with a proven body is how copywriters have won since then. - **Research the prospect's beliefs, not just the market.** Same destination, different starting point means a different direction if the prospect stands somewhere else, the route you give them points them away from buying. You cannot break a belief built over months inside one video, so map what they already believe and channel those beliefs toward your mechanism instead of arguing with them. - **Sell first, build later.** He validates by selling a blank deliverable (a placeholder PDF) and refunding if it does not convert, because the market vote is the only proof that matters. Start from the offer: list every objection, then design bonuses specifically to kill each objection, and model deliverables that have already been paid for rather than reinventing the product. - **One mechanism, 1,500 2,500 words, built on a fixed formula.** Match the word count of the VSLs winning in your market, then follow the line: there is this problem or opportunity, most people handle it wrong, here is the new way that is working, look at Pedrinho and Joaozinho plus testimonials. The craft rule is Gary Halbert and John Carlton's: say everything needed to make them believe in the fewest words possible. - **Demonstrate, never assert.** Show the mechanism working the cream applied and the result visible, the supplement explained at the point of action because the worst copy in the Brazilian market is the "believe me, I am the source" claim with zero proof behind it. Back claims with specifics (the Harvard-study pattern), and use analogies that make an invisible benefit physical, like sweeping your arteries clean or a mechanic diagnosing a broken VSL. - **Test the creative before you write the lead.** Spin up a throwaway page with a random video and a tiny budget purely to read CTR on competing hook angles promise, secret, conspiracy, mechanism. By his statistics the creative is the decisive variable: a good one works 8 times out of 10 uploads, a bad one 1 out of 10, so the winning CTR angle becomes the lead and its headline; an alternative team validates with initiate checkout instead of CTR. - **Run two ad approaches and know what each buys you.** The congruent approach has the ad talk about the mechanism itself "I will teach you a new way to get this benefit" carried unchanged through ad, lead, and VSL; it is the easy fruit but has few chances to hit. The indirect approach hooks an adjacent topic (Botox, why some women produce more collagen) and converges on the mechanism later; it reaches the far larger audience that is not yet aware of your mechanism. Peter Kell's trick is the template: hook with a random tip (rice is aging you), explain why, tie it to the problem mechanism, and the solution lands before they saw it coming. At Thiago Filemon's creative immersion he also saw hook stacking three unrelated hooks in one validated creative body going out converting. - **Ship the ugly VSL to buy data.** His quick validation VSL is 10 12 minutes on a white PowerPoint background; one such video turned R$2,000 of spend into R$7,000 of sales. He also A/B's a polished VSL against a one-day PowerPoint version: the good cut might run ROAS 3, the ugly one 1.5, which tells you the editing gain you will unlock after the offer itself is proven. - **Review the copy for its emotional arc, not just its logic.** Read the whole letter asking what the prospect should feel at each point hope at the top so they keep reading, agitation in the middle so they feel the problem, hope again when the solution arrives (the pattern he credits to Jason Fadl's "One to Many"). His course example: pitbull attack statistics persuade nobody, the story of parents finding their child's scalp torn open does and the statistics placed right after the story become proof instead of noise. ## 3. Methods & Setup - Step-by-Step 1. Open a dedicated research inbox a separate email address used only to receive competitor swipe and funnel emails; setup: one fresh inbox per business, subscribe to every US offer in your niche, and log each offer's sending domain so you can pull the full sequence later. 2. Mine the five most-scaled cold-traffic ads in your market copy their promise, pain, and mechanism into one swipe document; setup: spy tools plus YouTube keyword search (he teases a tool that searches ads by their transcript), note which angle each scaling ad runs. 3. Unmask who is behind the scaling offers search the domain's company registration to find the operation running it; setup: Google search of the domain's CNPJ/registration data, record operator, offer, and traffic source per finding. 4. Build the prospect's belief inventory list what he already believes about the problem and where he stands today; setup: a one-page belief map with the "same destination, different starting point" test applied to every belief you plan to use. 5. Write the offer skeleton before the product objections first, bonuses engineered to kill each one; setup: objection list, one bonus per objection, and a placeholder deliverable (blank PDF) with a refund clause for the validation sale. 6. Choose the single mechanism and adapt it across niches the same mechanism can sell in an adjacent market if the belief transfer holds; setup: mechanism statement in one sentence, plus the two or three markets where the underlying belief already exists. 7. Draft the mechanism at 1,500 2,500 words following the formula problem, wrong way, new way working, named proof, testimonials; setup: word count matched to the winning VSLs you swiped, Halbert's "fewest words to make them believe" as the trim rule. 8. Run a demonstration pass replace every "trust me" assertion with the mechanism shown in action; setup: cream demo, supplement demo, or Harvard-style specific, plus one analogy that makes the benefit physical. 9. Pick the ad approach per angle congruent (ad talks mechanism) or indirect (adjacent hook that converges); setup: one congruent creative and two indirect ones from the hook bank (promise, secret, conspiracy), all pointing at the same lead. 10. Pre-test creatives before writing any lead a throwaway page with a random video, smallest budget that returns statistically usable CTR; setup: three to five hook variants, CTR as the read (initiate checkout as the alternative if you can place a mini sales video), winner becomes the lead's angle. 11. Write the lead as the literal continuation of the winning ad same idea, same emotional entry state; setup: ad hook sentence carried into the first lines of the lead and again into the VSL opener, so the congruence chain ad lead mechanism never breaks. 12. Model the winning VSL's structure word count, opening, where the testimonial sits, what follows it; setup: your outline mirrors a paid-for structure (the John Caples "Do you make these mistakes in English" pattern of recombining proven headlines), no invented format. 13. Ship the ugly PowerPoint VSL 10 12 minutes, white background, no production value; setup: smallest ad budget that returns CPA, compare ROAS of the ugly cut against the polished cut to price the future editing gain. 14. Do the emotional-arc revision pass hope, agitation, hope; setup: read the full letter once asking "how does this make him feel here", place the statistics immediately after the story that earns them. ## 4. Tools & Stack - what each one is for here | Tool | Role | In this episode | | --- | --- | --- | | Spy tools (AdMiner-type) | Find the ads actually scaling in a niche | Source for the five most-scaled cold-traffic ads he researches before writing | | YouTube keyword + transcript search | Locate hooks and angles by spoken words | His teased "search ads by transcript" method for finding what is running | | ClickBank | Survey validated US offers and payouts | One of the marketplaces he checks to see what is already selling | | SemRush | Demand and keyword research around a niche | Listed among his market-research stack | | Google CNPJ/registration search | Unmask the company behind a scaling domain | The trick to see who is operating an offer you found | | Dedicated swipe inbox + offer domains | Receive and segregate competitor funnels | Separate inbox that only receives US swipe; domains logged per offer | | PowerPoint | Ultra-fast VSL production for validation | The 10 12 minute white-background VSL that bought his R$2,000 R$7,000 data | | Facebook Ads (min-budget CTR test) | Validate a hook before any lead is written | CTR read on cheap throwaway pages; creative is his 8/10 vs 1/10 variable | ## 5. Actionable Checklist - [ ] Create a dedicated research inbox and subscribe to every scaling US offer in your niche. - [ ] Swipe the five most-scaled cold-traffic ads; extract promise, pain, and mechanism from each. - [ ] Unmask the operators behind the best ads with a domain registration/CNPJ search. - [ ] Write the prospect's belief map and mark which beliefs you will ride instead of fight. - [ ] List every objection, then design one bonus per objection before building the product. - [ ] State your single mechanism in one sentence and name two markets where the belief transfers. - [ ] Draft the mechanism at the market's word count (1,500 2,500) using the problem / wrong way / new way / proof formula. - [ ] Replace every unsupported claim with a demonstration, a specific, and one physical analogy. - [ ] Build one congruent and two indirect hook angles, all converging on the same mechanism. - [ ] Pre-test the hooks on a throwaway page; pick the winner by CTR (or initiate checkout) before writing the lead. - [ ] Write the lead as the exact continuation of the winning ad, then model the winning VSL's structure. - [ ] Ship the ugly 10 12 minute PowerPoint VSL, buy CPA data, and only polish after the offer proves out.