RAW: the road to $2,000 a day at 2+ ROAS

Nov 3, 2026. Built from: our last 90 days of ads (204 ads, $16,943), today's research on GODA, Penrose and Cupids (1,170 ads, 773 transcripts), our Meta account playbook, and a full plan from GPT-6 Astra (high), which I corrected where it went soft. Nothing has been changed in the account.

1. The math we have to beat

• $2,000 a day at 2.0 = about 80 orders a day, about 120 bottles a day. Stock for a month is roughly 3,500 bottles.
• Two ways to close the gap: cheaper sales (better ads) and bigger orders (more 3-packs, a holiday bundle). We need both. Getting 3-pack share to 45% lifts the order to about $56, which moves the line from $25 to $28.
• Our own numbers: statics made 366 of 389 sales at about $41 each. Videos made 23 sales at about $90 each. Videos got lots of clicks but didn't sell.

2. Why we're stuck (biggest first)

1. One winning idea. Almost all our profitable sales came from one ad family (the GODA-style static). GODA has about 24 live story ideas in a single week's drop; we have one.
2. Too few new ads, too slowly. We launch a couple a week and judge each alone. They launch 25 to 90 a day and let Meta find the winners.
3. Too many hands and edits. The account gets changed every 1 to 3 days, which keeps resetting Meta's learning. Tests launched on the same men have cut our winners' sales every time.
4. Videos built as spectacle, not sales. The UFO and reaction videos got clicks but didn't show the bottle, the reason or the price early enough.
5. Small order size. At $50 average there's no room for a mediocre ad.

3. How to run the account

Now ($120 to $300 a day)

• Keep every current winner exactly where it is. GODA-style static in its age groups, the apology ad, the retargeting ad (it's one of our best per sale). No moving, no copying into new ad sets.
• Add one "drop" campaign for new ads, budget set at the campaign level, capped at about 20 to 25% of total spend (about $30 to $40 a day now). Same men 25 to 64, same placements as the winners, so the ad is the only thing that changes.
• Each week, one new ad set ("drop") with 8 to 10 new ads: 3 ideas, about 3 cuts each (different first line or opener). This is the competitor method: put a batch in, let Meta push money to the 1 or 2 it likes, and ignore the rest.
• Don't pause a whole drop because it's slow. Pause one ad only when it has spent about $75 with no sale. Ads that never got spend aren't losers, just unread.
• A winner stays in the drop it was born in. If a drop holds a proven ad (10+ sales in 7+ days at $28 or less), that drop's budget grows. It is never copied somewhere new. That rule already cost us sales once.
• Drops with nothing after 7 days are switched off. That keeps tests at 20%.

At $500 a day

• Winners about $400, drops about $100 (2 drops a week).
• Run one fair test of open targeting (Meta picks the men, no age walls) using our best ad, $100 a day for 7 days. If it sells as well as the age groups, move money over slowly. If not, keep the age groups. We built them for a real reason; this is the first clean check of whether they still matter.

At $2,000 a day

• Two campaigns. "Scale": every drop that holds a proven winner, budget at campaign level, about 80%. "Drops": new batches, about 20% ($300 to $400 a day), 3 to 4 drops a week.
• Retargeting stays small (10% or less) and is judged on real store orders.
• This matches what GODA most likely does: batches in, nothing paused, Meta feeds winners. We just can't afford their 4,000 ads, so we keep the kill line.

4. How much content, and what mix

• "Endless content" helps only through drops. Make as much as you want, but launch on the weekly schedule. Extra ads wait in the queue; a backlog is good.
• Where the ideas come from: 60% new versions of our proven families (GODA-style, apology, guarantee dare, "Smelling nice never got anyone laid"), 30% competitor formats that have run 90+ days, 10% wild swings.
• Videos only move up the mix when they earn it: two video ideas each with 10+ sales at our target. Competitors' video volume is a reason to test videos, not proof they'll sell for us.

5. Videos to make (best bets first)

1. Warning / skeptic talking head (20 to 35 s). Man or woman to camera, bottle in hand in the first 2 seconds. "I thought pheromones were BS..." / "Do NOT wear this to work." Cupids and Penrose run this all year.
2. Reverse apology (20 to 30 s). Our apology static is a proven winner; this is the same idea talking. Penrose ran "We owe an apology... because we lied" 99 days.
3. Women's POV / ranking (25 to 40 s). "Top 3 scents that make her want to kiss you before she knows your name" with RAW at #1. GODA and Penrose both run long-lived versions.
4. Bar or elevator test (25 to 40 s). "Come with me as a 30-year-old to see if this works at the bar." Penrose, 115 days live.
5. Black Friday warehouse / overstock (20 to 30 s). "We're drowning in inventory" (Cupids, 147 days). Seasonal, right now.
6. AI cartoon story (35 to 50 s). GODA's main format and a top format across beauty and health. Small test, 2 ideas x 3 cuts. The bottle has to appear by second 5.
7. Podcast clip (20 to 35 s). Woman on a mic: "Most men don't know the hug decides it." Penrose and GODA both run these.

6. Statics to make (best bets first)

1. GODA-style family: big claim + bottle + both prices. New versions change one thing: headline, photo or proof line.
2. Apology / warning: "We should have warned you." Our second-best family.
3. Real review screenshot + bottle: e.g. "She kept smelling my neck for 10 seconds."
4. Swearing manifesto: Penrose's "F*CK your cheap cologne" style, text-heavy.
5. Designer price trap: $300 bottle, zero compliments, vs $39.90.
6. Guarantee dare: our retargeting winner.
7. 3-pack math / Black Friday.
8. Cartoon-illustrated scene: Penrose now runs these as statics.

7. Angles (the reason to buy)

8. Video script ideas (hook = the first 2 seconds)

Starting points for the creative thread. Each still goes through our hook writer, the copy reads and your OK before anything is made.

1. "We should have warned you." Apology. Customer complaints that are really flexes, bottle reveal, why pheromones work, 3 for $75.
2. "Smelling nice never got anyone laid." Talking head. Your cologne is for you; this is for her; how pheromones hit before she knows why; review; offer.
3. "Do NOT wear this to the office." Warning. Coworker story, bottle, the science line, guarantee, offer.
4. "I thought pheromones were bullsh*t." Skeptic. What he tried, first night out, what she said, offer.
5. "Your $300 cologne is gone by lunch." Designer trap. Sniff test at 1 pm, RAW on the skin, price math.
6. "Come with me to see if this works at a bar." Live test. Spray, walk in, two moments, verdict, offer.
7. "Top 3 scents that make her lean in before she knows your name." Woman ranks 3; #1 is RAW; why; offer.
8. "She smelled my neck for ten seconds and didn't say a word." Review story told by the man. Bottle, mechanism, offer.
9. "My wife told me to stop wearing this out without her." Couple. Jealousy as proof, bottle, 3-pack.
10. "Married 12 years. She stopped reaching for me." AI cartoon story. The night it changed, the bottle, the turn, offer.
11. "Divorced at 44 and nobody looked at me." AI cartoon or photoreal story. First night out after, bottle, result, offer.
12. "We are drowning in inventory." Warehouse Black Friday. Pallets, deal, why it works, deadline.
13. "I feel bad for the guys who paid full price yesterday." Black Friday regret. Deal, bottle, deadline.
14. "Ladies, if your man smells like nothing, send him this." Woman to camera. Bottle, what changes, offer.
15. "Most men don't know the hug decides it." Podcast clip. Why scent decides the hug, RAW, offer.
16. "This didn't work. This didn't work. This did." Bathroom shelf of dead colognes, RAW last, offer.
17. "Bro, just get Baccarat." Two-friend skit. One spends $300, one spends $39.90 and gets the attention.
18. "The bartender walked past three guys to get to him." Photoreal scene. Reveal what he's wearing, offer.
19. "If you're shy, don't buy this." Warning talking head. What happens, why, guarantee.
20. "Women begged us to make it weaker." Founder line. Formula story, bottle, offer.
21. "Reply to @comment: 'Pheromone cologne is a scam.'" Comment reply. Show the bottle, review, guarantee.
22. "What does she actually smell when she hugs you?" Woman's POV. Sweat, laundry, cheap cologne vs RAW.
23. "He got three numbers and he's not even good-looking." Friend skit at a bar. Reveal, offer.
24. "The real reason she stopped texting back." Story. First date, the hug, the miss, the fix.
25. "Don't buy this unless you want to be the guy she can't stop thinking about." Warning close. Bottle, offer.

9. Static ideas (headline / picture / offer)

1. "WE SHOULD HAVE WARNED YOU." / bottle + 3 "complaint" reviews / $39.90 or 3 for $75
2. "SMELLING NICE NEVER GOT ANYONE LAID." / bottle on a bar / both prices
3. "F*CK YOUR $300 COLOGNE." / designer bottle in a trash can, RAW standing / $39.90
4. "SHE SMELLED MY NECK FOR 10 SECONDS." / review screenshot + bottle / 3 for $75
5. "DO NOT WEAR THIS TO WORK." / office, bottle on desk / $39.90
6. "YOUR COLOGNE QUIT AT LUNCH." / clock at 1 pm, bottle / both prices
7. "I THOUGHT PHEROMONES WERE BULLSH*T." / handwritten note + bottle / guarantee
8. "SHE LOOKED RIGHT PAST YOU." / bar scene, man blurred / $39.90
9. "MARRIED 12 YEARS. SHE STARTED REACHING AGAIN." / bedroom lamp, bottle / 3-pack
10. "LADIES: HE NEEDS THIS MORE THAN ANOTHER TIE." / gift box / Black Friday 3-pack
11. "3 BOTTLES. $75. DO THE MATH." / three bottles / save $44.70
12. "BLACK FRIDAY: WE OVERORDERED." / pallets / deal + deadline
13. "WORKS OR YOUR MONEY BACK. WE DARE YOU." / bottle + guarantee seal / $39.90
14. "THE GUY SHE CAN'T STOP THINKING ABOUT ISN'T BETTER LOOKING." / two men / $39.90
15. "DIVORCED AT 44. INVISIBLE AT 45. NOT ANYMORE." / mirror shot / $39.90
16. "THIS ISN'T COLOGNE. IT'S AN UNFAIR ADVANTAGE." / bottle hero / 3 for $75
17. "WHAT SHE SMELLS WHEN SHE HUGS YOU." / hug, close / both prices
18. "STOP SPRAYING. START GETTING NOTICED." / bottle / $39.90
19. "HE'S WEARING $40. YOU'RE WEARING $300." / two men at a bar / $39.90
20. Cartoon static: "CAN'T SMELL HIS COLOGNE BY DINNER?" / illustrated date / both prices

10. Offer, order size, new products

• Black Friday now: all three competitors are in holiday mode. Test "Buy 2, get 1 free" at $79.80 for 3 against the current 3 for $75. It's more money per order and reads as a bigger deal. Put a deadline on it.
• Make the 3-pack the obvious choice: show "$25 a bottle, save $44.70" next to the buy button, with reviews right there. Goal: 3-pack share from 29% to 45%.
• Free gift test (cheap travel spray or sample) only if it raises money per visitor.
• New products (after the cologne works): GODA and Cupids both just moved into men's pheromone body butter, Penrose's whole business. A RAW pheromone body butter or oil is the obvious second product. New products give us new ad libraries to fill. Gift set first, body butter second.

11. Decision rules

• Daily: spend, store orders, store ROAS, cost per sale, average order, 3-pack share.
• Ad kill line: about $75 spent with no sale (3x our $25 line), after a day's lag. Below $28 spent = not read yet.
• Winner: 10+ sales over 7+ days at $28 or less.
• Budget steps: +15 to 20% no more often than every 3 days, only when 7-day store ROAS is above 2.2 and 14-day above 2.0. If the next step gets worse, go back.
• One decision owner. Everyone else proposes changes; changes ship on a schedule: drops Monday (and Thursday at $500+), budget moves every 3 days.
• Judge money on Shopify, not Meta. Meta decides which ad gets credit.

12. 90 days, with gates

Where I disagreed with Codex

• Its hooks were soft ("Smell like you made an effort"). Our winners and every competitor's long-runner sell attraction hard. I replaced them with competitor-proven lines pushed further.
• It said to switch off retargeting. That ad is one of our best per sale, so I kept it small and judged on store orders.
• It said to add new ads into the existing winners' ad sets. That has cut our winners' sales before, so new ads go into separate drops.

Sources: research/growth-plan-2026-11-03 (Astra plan + prompt), research/competitor-copying-2026-11-03, research/competitor-check-2026-11-03, 90-day ad export.