quick one on testing and scaling static image ads for fashion dropshipping. no video-in this niche, video takes forever to convert compared to a solid image.
Scenario 1: Find a product, bang out 6 static creatives into a single broad ad set inside a CBO campaign at $50/day. day 1, three ads show promise, one dominates. day 2, same ad pulls 2-3 sales. day 3, another couple of orders. Now how the hell do I scale from here?
Double budgets? increase 20-30%? duplicate the ad set? go horizontal? My issue is the ad set needs to exit learning properly-once it hits Learning Limited, performance tanks hard. I know people say Limited can still perform, but Meta themselves tell you to adjust settings when it hits that status, so it clearly matters in 2026. aggressive scaling like doubling gets you to 50 conversions faster, but I'm trying to see what actually works for others right now.
Scenario 2 (my current mess): Same campaign out of learning, running $200/day, one winning image ad bringing 15-20 purchases consistently. Then this week-Tuesday onward-performance drops hard. Already cut budget by 20%, still recovering. ATCs are high, CTR's decent, but conversions evaporated.
so where do I test new iterations of the same winning angle? option 1: add new ad sets into the existing winning CBO campaign. Option 2: leave that campaign untouched, create a fresh CBO, test there separately. I'm leaning toward Option 2 to protect the margin, but curious what's actually moving the needle for people in fashion image ads right now