1. Consolidate the Account Most accounts I audit are running a dozen campaigns on $xx-a-day ad sets that will never exit learning. Meta needs roughly 50 optimization events per ad set per week to figure out who buys from you. Starve it and it guesses- which translates to you leaving money on the table.
The Move: Fewer campaigns, fewer ad sets, more creative variety inside each one. How do you decide what campaigns to keep? Start with KPI guardrails for sunsetting campaigns! Too often we get married to campaigns. Data>feelings.
The Payoff: I’ve watched brands go from 15 campaigns down to 3 and see overall ROAS climb 30%. Cleaner structure, cleaner signal.
2. Cap Your Existing Customers Your sales campaigns will happily “prospect” people who were already coming back, then hand you inflated ROAS and take the credit.
The Move: Set the existing customer cap to 20–30% on your scaling campaigns.
The Why: You’re paying acquisition prices for retention traffic. Cap it and your prospecting numbers become numbers you can trust.
3. Build a Creative Testing Lane Don’t test new concepts inside your scaling campaign. You’ll pollute the winners and starve the tests.
The Setup: One small ABO (ad set budget) campaign where new angles prove themselves. Winners graduate to your sales campaigns.
The Cadence: Refresh creative every 25–35 days. CPM up 30% from baseline, CTR under 1%, frequency over 4 =fatigue, not a market shift. Pay close attention to these numbers
4. Fix the Signal Before You Blame the Algorithm Meta can only find buyers that look like the buyers it can actually see. Blurry signal in=expensive guesses out.
The Setup: Pixel plus Conversions API, with an Event Match Quality score of 6 or higher. Server-side tracking if you’re serious / ad blockers and browser restrictions are eating your client-side events alive.
The Rule: If your event data is distorted, your CPA is fiction. Fix tracking before you touch targeting.
5. Scale Smart- Doubling budget overnight resets learning and torches your CPA. Every time.
The Move: Increase budgets 20% every 3–4 days. No single swing bigger than 20%.
The Math: Minimum viable daily budget per ad set is roughly your target CPA × 50 ÷ 7. Below that, you’re not giving the algorithm enough conversions to learn from.
If you’re about to raise budget on an account with a shaky foundation, you’re just paying Meta to find out faster. I run fractional growth audits to pinpoint these exact leaks and hand you the execution hitlist. Let’s fix the foundation before you scale it. Connect