## 1. Executive Summary

Thiago Lucena has been buying paid traffic since 2020 and now runs affiliate campaigns for physical products in the US market, averaging about $1M per month (R$5M) roughly 10x the R$500-600k/month he reported at a mentorship event earlier in his career. His whole doctrine is one line: respect the conversion. A converting offer is a rare money machine that can stop tomorrow, so you extract it at full speed, reinvest everything it returns, and never slow down to buy status. Today he runs Facebook as the main source, outsources account contingency, and credits his consistency to spying, structured A/B tests, and slow analytical decisions instead of gut calls.

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

- **Respect the conversion and squeeze the winner before it dies.** On his first hit offer he did not respect it and made R$1M; on the second he did, and made R$4M, almost R$5M combined. When an offer converts, do not buy a car take every spare dollar back into the campaign, because the machine is not infinite and you do not know when it stops.

- **Spy first: the 80/20 of being an affiliate is espionage.** Watch the offer climb its network ranking week over week, check page traffic with SimilarWeb to see if it is being scaled, then use AdSpy to see who is running it and on which traffic source. Take the producer's checkout into SEMrush, pull the domains sending traffic to it, and hunt those affiliate domains in the Facebook Ad Library that gives you the angles, funnel shape, and traffic source before you spend a cent.

- **Buy in the US auction where commissions are 70-75%.** US producers pay 70-75% (sometimes 75-80%) because the affiliate is doing the acquisition; average ticket is around $200 versus roughly R$250 for a PLR ebook, so a CPA of R$300-400 can still be profitable and you outspend the ebook crowd in the auction. Upsells are 30-35% of revenue and maybe 60% of profit one tracking failure on the upsell cost him an estimated R$10,000 in two hours.

- **Kill the ad, not the offer.** What dies first is the creative, not the funnel people tire of the parts they see most, which are the openings: the start of the creative, then the start of the VSL. Keep three lead types running (scientific discovery, promise, story), keep the ad angle congruent with the lead, validate the ad before you touch the VSL, and keep the mechanism untouched. Long copy wins: when he compressed a 25-minute VSL to 20 minutes, it converted worse.

- **Lateralize accounts instead of inflating one.** Run R$10,000-20,000 per account across many accounts, because Facebook spreads the buying audience across advertisers and one big account only gets a slice of it. Accounts behave like offers roughly 1 in 4 (20-25%) pops so continuity comes from volume. Contingency is outsourced to a firm on a fixed fee (about R$3,000) plus a percentage of spend (example given: 5% on R$100,000 = R$5,000), which buys accounts that already carry clean spend history and score.

- **One fixed campaign skeleton, one budget ladder.** One campaign = 10 ad sets x 3 ads; in validation the 3 ads are totally different, in optimization it is one winning creative plus 2 hook variations; clone it 10x for 100 ad sets and 300 ads. Open at R$200 at US midnight (5 a.m. local), lift to R$500 once the day cycle starts never open at R$500, the account needs a full 24-hour cycle to learn peak hours. Scale rule: ROAS above 2, double the budget; below 2, cut it in half (his ladder ran 150 300 600 1200; at 1200 it returned 1.8, so he cut to 600).

- **Be white in the ad, black after the click.** Keep promises indirect in the creative story-driven, mini-VSL style, no absurd claims so approvals and accounts last, and put the real aggression in the VSL where the click already happened. It extends campaign life: an over-black ad can wake up to everything rejected. The same logic applied on YouTube, where homegrown screen-recorded phone ads spent better than polished ones, which the algorithm dumped into display.

- **Stitch a "super VSL" from tested parts, and run accounts in dollars.** Push every candidate VSL through your player, read retention plus conversion, then assemble the best lead, the best mechanism, and the closing that actually sells 6-12 jars, narrated by one voice. One graft took an offer that was barely selling into scaling the same week. Moving ad accounts to US-dollar cards improved delivery and score: his VSL conversion went from 1.34 (a two-month metric) to 1.70 in a single month, with the Black Friday seasonality caveat noted.

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

1. **Pick the network in priority order.**; setup: BuyGoods first (hardest to enter, best checkout approval), then ClickBank (low barrier, good support for finding products), then Digistore (low barrier, but the best offers stay hidden until you have results and an affiliate manager). Confirm the payout method works for you before sending a dollar no payout rail, no platform.

2. **Find offers that are entering hype, not offers already in it.**; setup: watch network ranking week over week, confirm rising page traffic with SimilarWeb, then check AdSpy for who is scaling it and on which traffic source. An offer on page five that gets steady daily sales is more interesting than a flat top offer.

3. **Reverse-engineer the top affiliates.**; setup: producer checkout SEMrush affiliate domains their ads map the funnel (bridge, pre-bridge, VSL, upsells). Expect cloaking and spoofed redirects break them with referer/cookie tricks or a clean VPN before you judge the funnel.

4. **Differentiate instead of copying.**; setup: keep the skeleton and the mechanism, change the voiceover, the photography, and rewrite the hook. Exact copies get deprioritized by Facebook against the original uploader, so give it a new face without changing the essence.

5. **Build the campaign skeleton.**; setup: 10 ad sets x 3 ads, cloned 10x (100 ad sets, 300 ads); R$200 opening budget at US midnight, lifted to R$500 around 5 a.m. local; validation = 3 distinct ads, optimization = 1 creative + 2 hook variations; cap validation at up to 5 accounts.

6. **Lateralize and buy contingency instead of DIY logins.**; setup: R$10,000-20,000 per account across many accounts; contingency vendor on a fixed fee (~R$3,000) plus a percentage of spend (~5% example), with overnight and weekend coverage, because that is when the worst problems land.

7. **Optimize day-over-day, never intraday.**; setup: do not touch campaigns midday a day that starts bad can flip profitable by evening cut budget on yesterday's losers, leave winners alone, then re-upload everything fresh the next morning as if the previous day never happened. Renew every active creative before the weekend: midweek enforcement sweeps dropped them from $40-50k days to $20k days, which wrecks the monthly average.

8. **Revive a dead offer at the creative layer first.**; setup: swap the openings (creative start, then VSL lead), keep the mechanism, rewrite the background story from books written by people with the problem, and keep all three leads discovery, promise, story congruent with the ad angle. Validate the new ad first, then move the VSL lead.

9. **Assemble the super VSL.**; setup: test all candidate VSLs in your player, keep the highest-retention lead, the highest-converting mechanism, and the closing from the offer that sold 6-12 jars at $200 commission; one narrator throughout, copy hand-finished so the pieces do not read as disjointed.

10. **Switch accounts to US dollars.**; setup: dollar cards on a US-timezone account profile improve score and delivery in the US/Canada/UK/Australia/New Zealand auctions; watch for seasonality before crediting the whole gain to the change.

11. **Staff the analysis, not the uploading.**; setup: three media buyers at the start (Thiago, his partner, and their lead buyer Glen), all from the same mentorship; uploading campaigns is robotic, analysis and decisions are the 80/20 and reserve one noon-1pm block a week in the calendar purely to decide the next highest-leverage move.

12. **If you are starting today with almost nothing, skip the VSL work.**; setup: run bottom- and mid-funnel search ads (Google Ads / Bing Ads) as an affiliate for offers in the US market pure text ads, no account bans, no video and use it to accumulate cash before hiring editors and copy.

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

| Tool | Role | In this episode |
|---|---|---|
| Vurb | Video hosting, retention and A/B testing | Their new test lets you add videos mid-test and split traffic by percentage; an autoplay layout swap lifted conversion ~50% for a user |
| Facebook Ads | Primary traffic source today | R$10,000-20,000 per account, lateralized across many accounts with outsourced backup |
| YouTube Ads | Former main source, still viable | Where he first scaled: one campaign returned R$250,000 on R$100,000 invested |
| BuyGoods / ClickBank / Digistore | Offer networks | His stated priority order; BuyGoods for checkout approval, Digistore for affiliate analytics and promo-code tracking |
| AdSpy + SEMrush + SimilarWeb | Spy stack | Offer momentum, page traffic, and mapping affiliate funnels starting from the producer's checkout |
| Hotmart | Former producer platform | The chargeback block that pushed him out of producing and into the affiliate market |
| Incogniton + proxies | Account contingency | Fingerprint/profile separation after Facebook wiped his accounts in early 2021 |
| ChatGPT | Copy and code | Rewrites hooks and generated the PHP for his own traffic-filter (cloaker) |

## 5. Actionable Checklist

- [ ] Register on BuyGoods first, then ClickBank, then Digistore and verify the payout method before sending traffic.
- [ ] Track 10 offers in your niche weekly and flag anyone whose network rank climbs two weeks in a row.
- [ ] Run every rising offer's checkout through SEMrush and collect the affiliate domains sending it traffic.
- [ ] Build the 10 ad sets x 3 ads skeleton, clone it 10x, and cap your validation at 5 accounts.
- [ ] Open at R$200 at US midnight, lift to R$500 at 5 a.m., double the budget on ROAS > 2 and halve it on ROAS < 2.
- [ ] Move to R$10,000-20,000 per account across multiple accounts and buy outsourced contingency.
- [ ] Keep three leads ready discovery, promise, story and match each one to the ad angle running against it.
- [ ] Renew every active creative before the weekend and never touch campaigns midday.
- [ ] Keep the promise indirect in the ad and put the hard sell after the click, inside the VSL.
- [ ] Reserve one noon-1pm block per week to decide the single highest-leverage move in the business.