The Creative Budget Allocation: What % to Testing — Advanced Facebook Ad Creative Testing

Quick answer

Allocate 10-15% of total Meta spend to creative testing. Below 10% you starve the testing pool and run out of winners. Above 15% you starve the evergreen and miss your CPA target. The exact number depends on volatility — newer accounts and volatile categories trend toward 15%; mature stable accounts toward 10%.

Why a fixed % beats vibes

The "I'll test when I have time" approach guarantees that testing never happens. The "I'll test until I find a winner" approach guarantees over- or under-spending depending on luck. A fixed %, ringfenced from total spend, makes testing a permanent line item that survives every internal pressure.

The framework, step by step

1. Calculate total monthly Meta spend.
2. Reserve 10-15% as the creative testing budget. Lock it in the campaign structure as a separate testing CBO.
3. Inside the testing CBO, run weekly cycles: 5-15 new creatives per week.
4. Apply the £50 threshold rule and 4-day duration rule per test.
5. Promote winners to evergreen weekly. Demote losers.
6. Audit the % monthly. Adjust if winners are stale (raise %) or evergreen is hitting CPA (lower %).

Example test matrix

Monthly spend	Testing %	Testing budget	Approx new ads/week
£3,000	15%	£450	5-8
£10,000	12%	£1,200	15-25
£30,000	10%	£3,000	30-50
£100,000	10%	£10,000	60-100

Pitfalls to avoid

• Mixing testing budget with evergreen budget. The auction always favours proven ads.
• Letting the testing budget creep below 10%. By month 3 you have no fresh winners.
• Spending the testing budget on new audiences instead of new creatives. Creative is the highest leverage.
• Treating testing as overhead. It is the R&D line item; cutting it kills the account.
• Testing only when CPA breaks. Test continuously to pre-empt the break.

5 FAQs

Does this include retargeting tests? No. Retargeting is a separate budget line.

What if I cannot afford 10%? Then your account is too small to scale predictably. Test on whatever you can afford and expect noisy results.

Should I increase % during a campaign launch? Yes — 20-25% in the first 4 weeks of a new product.

Can I cut the % when CPA is great? Yes, but only to 8% minimum. Below that, the bench dries up.

Should the % change with seasonality? Yes — increase before peak season, decrease during.

Make 10-15% go further

Pix-Vu helps you turn the testing budget into more variants per pound — same spend, more shots on goal. https://pix-vu.com.