September is the most important month of your BFCM creative strategy. This is when you generate, test, and identify the winning creative that will drive revenue in November. Every dollar spent on testing in September saves $5-10 in wasted spend during November.
Generating the test library
The goal for September: produce 30-50 creative variants and test them all at low budget. Use Mani's daily queue and on-demand generation to produce this volume. With a daily queue of 5 ads per day, you get 150 variants in September. You only need to test 30-50, so you have room to be selective about which ones to test.
Generate across three creative concept categories. Offer-focused: "30% off everything," "$20 off your first order," "Free shipping on BFCM." These test which offer type resonates with your audience. Product-focused: feature specific products or collections that will be promoted during BFCM. These test which products generate the most interest. Story-focused: founder narrative, customer testimonials, brand values. These test whether emotional connection drives better results than promotional messaging.
Within each category, generate variants with different hooks (question, statistic, bold claim, UGC-style), different formats (static, video, carousel), and different CTAs (Shop now, Save now, Get early access). The goal is diversity. You are exploring the creative space to find what resonates.
Running the tests
Launch all 30-50 variants in a single Advantage+ campaign on Meta with a modest daily budget ($50-100 per day total). Advantage+ will automatically distribute budget to the top performers. Let the tests run for 7-10 days to accumulate meaningful data.
On TikTok, create a similar test campaign using Smart Performance Campaign. TikTok's algorithm needs more data to optimize, so plan for 10-14 days of testing.
Do not judge individual ads before 48 hours and 1,000+ impressions. Early performance is noisy. Wait for the data to stabilize before making decisions.
Identifying winners
After the testing period, rank all variants by CPA (cost per purchase or cost per lead, depending on your goal). The top 10% are your winners. The top 25% are your strong performers. Everything else gets paused.
Analyze the winners for patterns. What do the top 5 ads have in common? Same hook type? Same format? Same product? Same offer? These patterns become your BFCM creative playbook. They tell you exactly what to produce more of in October's scaling phase.
Document the patterns in a simple brief: "BFCM 2026 creative winners: question hooks about pricing outperform statement hooks. Video outperforms static by 2x. The Hero Product (name it) generates 40% lower CPA than the full catalog approach. Urgency CTAs ('Last chance') outperform standard CTAs ('Shop now') by 25%."
This brief becomes your generation prompt for October. Feed these insights into Mani's guided generation to produce more of what works. Continue to October: scaling your winners.