🎓 Apply for the Accelerator (live cohort coaching): https://start.agencyclub.co/accelerator
đź”— Everything I offer, free and paid: https://agencyclub.co/
⚡ 30-day trial of GoHighLevel: https://gohighlevel.com/samueldarby
Every guru selling you a Meta ads testing framework is lying to you. Not on purpose. They're just running ecom accounts with ÂŁ1,000 a day, where the pixel gets 50 conversions per ad set every week and testing actually works.
Your local roofer gives you ÂŁ50 a day. Best case, that's 15 leads a week across the whole account. Spread over five ads, each one gets three. You're not testing anything. You're guessing with extra steps.
I've run local Meta ads for eight years. I stopped testing completely, started spamming ad creatives instead, and results got better. This video shows you the maths behind why, how Andromeda actually picks your audience now, and the exact spam-and-replace system I run on every account: one CBO campaign, five ads minimum, a target CPL set before launch, pause anything at 150 to 200% of target, replace it the same day, stop touching it.
WHAT YOU'LL LEARN
• Why the 50-conversions-a-week rule makes split testing pointless on local budgets
• How Andromeda uses your creative as the targeting (and why one ad can run for 6 months)
• The multiplier: 3 ICPs × 10 hooks × 10 formats = 300 ad variants from one market
• The five rules: one CBO, launch big, target CPL first, the 150 to 200% cut line, replace same day
• The three windows to judge an ad on (30 day, 7 day, 48 to 72 hours) and when to bench one
• How to take a winning hook into five new formats without changing the angle
TIMESTAMPS
0:00 Every testing framework guru is lying to you
0:56 Why testing works for ecom and not for local
1:30 The ÂŁ50 a day roofer: 15 leads a week, not 50
2:13 Andromeda: your creative is the targeting
2:59 The bit about Andromeda everyone forgets
3:40 One ad can carry an account for six months
4:03 Market → ICPs → hooks → formats
5:06 10 to 15 formats for every single hook
5:31 300 ad variants from one market
5:46 Rule 1: one campaign, CBO
5:53 Rule 2: launch with five ads minimum
6:09 Why not 10 or 15 ads (the budget maths)
6:41 Rule 3: set your target CPL before launch
7:03 The conveyor belt: pause and replace
7:27 The 150 to 200% cut line
7:56 Local Ads Lab
8:59 The spam-and-replace diagram
9:55 Judge ads on 30 days, 7 days and 48 to 72 hours
10:25 When to bench an ad and bring it back
11:11 Turning one winning hook into new formats
12:11 Spam until it works, not test until it fails
12:41 Load the account from day one
12:58 Where to go next
WHO THIS IS FOR
Agency owners, freelancers and local business owners running Meta ads on real-world budgets for home improvement and local service businesses.
#metaads #facebookads #andromeda
🎓 Apply for the Accelerator (live cohort coaching): https://start.agencyclub.co/accelerator
đź”— Everything I offer, free and paid: https://agencyclub.co/
⚡ 30-day trial of GoHighLevel: https://gohighlevel.com/samueldarby
Every guru selling you a Meta ads testing framework is lying to you. Not on purpose. They're just running ecom accounts with ÂŁ1,000 a day, where the pixel gets 50 conversions per ad set every week and testing actually works.
Your local roofer gives you ÂŁ50 a day. Best case, that's 15 leads a week across the whole account. Spread over five ads, each one gets three. You're not testing anything. You're guessing with extra steps.
I've run local Meta ads for eight years. I stopped testing completely, started spamming ad creatives instead, and results got better. This video shows you the maths behind why, how Andromeda actually picks your audience now, and the exact spam-and-replace system I run on every account: one CBO campaign, five ads minimum, a target CPL set before launch, pause anything at 150 to 200% of target, replace it the same day, stop touching it.
WHAT YOU'LL LEARN
• Why the 50-conversions-a-week rule makes split testing pointless on local budgets
• How Andromeda uses your creative as the targeting (and why one ad can run for 6 months)
• The multiplier: 3 ICPs × 10 hooks × 10 formats = 300 ad variants from one market
• The five rules: one CBO, launch big, target CPL first, the 150 to 200% cut line, replace same day
• The three windows to judge an ad on (30 day, 7 day, 48 to 72 hours) and when to bench one
• How to take a winning hook into five new formats without changing the angle
TIMESTAMPS
0:00 Every testing framework guru is lying to you
0:56 Why testing works for ecom and not for local
1:30 The ÂŁ50 a day roofer: 15 leads a week, not 50
2:13 Andromeda: your creative is the targeting
2:59 The bit about Andromeda everyone forgets
3:40 One ad can carry an account for six months
4:03 Market → ICPs → hooks → formats
5:06 10 to 15 formats for every single hook
5:31 300 ad variants from one market
5:46 Rule 1: one campaign, CBO
5:53 Rule 2: launch with five ads minimum
6:09 Why not 10 or 15 ads (the budget maths)
6:41 Rule 3: set your target CPL before launch
7:03 The conveyor belt: pause and replace
7:27 The 150 to 200% cut line
7:56 Local Ads Lab
8:59 The spam-and-replace diagram
9:55 Judge ads on 30 days, 7 days and 48 to 72 hours
10:25 When to bench an ad and bring it back
11:11 Turning one winning hook into new formats
12:11 Spam until it works, not test until it fails
12:41 Load the account from day one
12:58 Where to go next
WHO THIS IS FOR
Agency owners, freelancers and local business owners running Meta ads on real-world budgets for home improvement and local service businesses.
#metaads #facebookads #andromeda