## 1. Executive Summary

Joao Tafuri is the offer and copy partner behind one of the most scaled infoproduct VSL offers in Brazil the "Singapore system" with more than R$40 million in 2026 alone and the 40MM-in-8-months run in the title. The episode is not a traffic tutorial. It is a post-mortem of what actually made the offer scale, told in the order it happened: attention first, then the offer, then the mechanism, and only then the VSL. His read on attention is blunt you are competing with the most sensational content on the feed (gossip, betting, the loudest creators), so the creative has to earn the stop in the audience's own register instead of assuming the product deserves it. The bigger lesson is the offer/product distinction: the audience kept asking about the product until the team drifted into explaining features, and people stopped buying. A system alone does not sell either the sellable thing is the "code" installed inside the system, a named mechanism the market already believes in, the way collagen was repositioned as "Egyptian glue" before being transposed onto the offer. He also shows the proof shortcut: instead of claiming a platform works with data, show sellers the audience already knows who made money on it. And he is explicit that the mechanism is optional in the VSL he withholds it in some builds and tests. Figures are operator-stated, not audited.

## 2. Alpha / Sauce - what actually moves the needle, and exactly how to run it

- **Attention is the first media buy.** You are competing with gossip, betting, and the loudest creators in the feed. If the opening does not earn the stop in the audience's register, the offer is never read. Design for attention before you design for persuasion.
- **The product is not the offer.** Their audience kept asking about the product until the team started explaining features instead of selling the transformation and people abandoned the product. Audit the VSL for product-talk versus offer-talk.
- **A system alone is not enough install a code.** The "Singapore system" is the system; the thing that sells is the code implemented inside it. Name the mechanism and make it the thing the buyer repeats.
- **Borrow names the market already believes.** Collagen sold as "Egyptian glue" works because the name carries the proof. Find the name your market already trusts, then transpose it onto your offer.
- **Authority by proxy beats data claims.** Instead of arguing that Shopee works with statistics, he showed Shopee sellers the audience already recognized. Same proof, borrowed trust.
- **Build familiarity, not novelty.** Attention blindness is real the prospect has seen the form before. Repeated names and formats create recall that a brand-new angle never gets.
- **Irresistible offer is a discipline, not a slogan.** The Leandro Ladeira lineage matters here: they over-indexed on launch execution and CPL and under-indexed on why the offer itself was irresistible.
- **The mechanism is optional in the VSL.** Tafuri deliberately withholds the mechanism in some VSLs. Test mechanism-led against transformation-led; do not assume the mechanism always wins.
- **100k 200k+ came from changing the VSL, not the traffic.** When the offer flattened, the fix was the VSL, not more spend.
- **Scale targets are not universal.** A 30k/day operator does not need the 300k/day build; the structure should match the target.

## 3. Methods & Setup - Step-by-Step

1. Map the attention landscape; setup: list the five loudest content patterns your buyer already watches, then write the opening in that register.
2. Separate product from offer; setup: mark every sentence in the VSL that explains a feature and rewrite it as a transformation or a proof point.
3. Define the system; setup: write the steps the buyer follows in one page this is the system, not the sell.
4. Install the code; setup: name the mechanism inside the system so it can be repeated in one sentence.
5. Borrow a name the market believes; setup: find the existing term or authority your buyer already trusts and attach the mechanism to it.
6. Prove with people, not statistics; setup: show recognized sellers or users who got the result instead of making a platform claim.
7. Build familiarity; setup: keep the winning names and formats in rotation long enough to be recognized, instead of replacing them every week.
8. Re-check the offer against the irresistible-offer standard; setup: write the promise, the proof, and the reason it is different in three lines.
9. Test mechanism-led versus transformation-led VSLs; setup: one variable, same traffic, same ticket.
10. When the VSL flattens, rebuild the VSL; setup: new narrative and structure before new spend.
11. Match the build to the target; setup: design the funnel for the daily number you actually want, not the biggest number you have seen.
12. Keep the offer at the center of the launch; setup: review launch assets for CPL/execution talk that has crowded out the offer.

## 4. Tools & Stack - what each one is for here

| Tool / Element | What it is for here |
| --- | --- |
| VSL | The asset that carries attention, offer, and mechanism in one sequence |
| The system | The step-by-step method the buyer follows |
| The code / mechanism | The named idea inside the system that makes the offer believable |
| Borrowed names | Existing market terms that already carry proof |
| Authority by proxy | Recognized sellers or users who got the result |
| Instagram | Attention and familiarity building |
| Shopee | The platform context used in the episode's proof examples |
| Launch / CPL operation | Distribution and execution necessary, but not the offer |

## 5. Actionable Checklist

- [ ] List the five loudest content patterns your buyer watches.
- [ ] Write the opening in that register before writing the offer.
- [ ] Mark every product-feature sentence in the VSL.
- [ ] Rewrite feature sentences as transformation or proof.
- [ ] Write the system on one page.
- [ ] Name the code inside the system in one sentence.
- [ ] Attach the mechanism to a name the market already believes.
- [ ] Replace one data claim with one recognized person who got the result.
- [ ] Keep winning names and formats in rotation for recall.
- [ ] Write the promise, proof, and difference in three lines.
- [ ] Test mechanism-led against transformation-led.
- [ ] Rebuild the VSL before adding spend when it flattens.