Creative Matrix

Overview

Generate a full set of ad creative concepts for any brand using the Creative Multiplication framework by Curtis Howland. The framework produces 27 unique creative concepts by combining 3 messaging angles (Pain Point, Desire, Social Proof) x 3 creative formats (Static, Video, Carousel) x 3 funnel stages (TOF, MOF, BOF). Each concept is distinct enough to avoid auction overlap in Meta's algorithm.

Attribution: The Creative Multiplication framework was created by Curtis Howland, based on $30M+ in Meta ad spend. This skill implements his methodology as published on LinkedIn.

Prerequisites — Asset Generation (Optional)

Brief/copy generation requires no additional tools. Asset generation is optional:

• Static image generation: Requires nano-banana skill (Gemini). Verify: python3 -c "from google import genai; print('ok')" and GEMINI_API_KEY in environment. Alternatively: fal.ai MCP server (Flux, SDXL)
• Transparent product cutouts (for compositing on ad backgrounds): Requires nano-banana's full 4-pass web-asset pipeline (generate_image.py → edit_image.py → make_transparent.py → pngquant). See nano-banana skill's references/web-asset-workflow.md
• Video generation: Requires fal.ai MCP server (Kling, MiniMax, Wan)
• No tools available: Output briefs only — the user or their production team handles asset creation

To detect availability, check if nano-banana scripts are accessible before offering asset generation:

python3 -c "from google import genai; print('nano-banana: ok')"

Workflow

Phase 1: Gather Brand Context

Before generating concepts, collect the following from the user. Ask concisely — do not overwhelm with questions. Prioritize the first three; the rest are helpful but optional.

1. Brand/Product — What is the product or service?
2. Target Audience — Who is the ideal customer? Demographics, psychographics, pain points.
3. Offer — What is the current offer? (discount, bundle, free trial, etc.)
4. Product Category Signal — Is the value immediately visible (fashion, jewelry, food) or does the customer need to believe a transformation (skincare, supplements, coaching)? This determines the static vs video weighting.
5. Existing Creative Winners — Any ads that have worked well? What angle/format were they?
6. Budget Tier — Starter (5-10 concepts), Growth (15-18 concepts), or Full Matrix (all 27).
7. Platform Notes — Stories-heavy? Feed-heavy? Reels? This affects format specs.

Phase 2: Generate the Creative Matrix

Load references/framework.md for the full framework reference including angles, formats, funnel stages, sub-triggers, benchmarks, and combination examples.

For each concept in the matrix, produce a structured brief:

## [Cell ID] — [Funnel] x [Format] x [Angle]

**Hook/Headline**: [The opening line or visual hook]
**Body Copy**: [Supporting copy, 2-3 sentences max]
**CTA**: [Call to action]
**Sub-Angle**: [Which sub-angle: frustration/fear/regret, aspiration/freedom/status, testimony/authority/relatability]
**Triggers**: [Psychological triggers activated]
**Format Spec**: [Dimensions, duration, card count, production notes]
**Targeting**: [Funnel stage targeting notes]
**Priority**: [High/Medium/Low based on rules below]

Prioritization Rules

Apply these rules when assigning priority and recommending test order:

• Allocate 80% of creative volume to TOF
• Weight statics higher for visually-obvious products (fashion, jewelry, food)
• Weight video higher for belief-driven products (skincare, supplements, transformations)
• Decision rule: SEE IT → static. BELIEVE IT → video.
• Carousels for education-heavy or comparison-selling products
• Mark Partnership/creator-handle ads as high priority when applicable (53% lower CAC)
• For starter budgets, recommend a balanced sample: 3 TOF (one per format), 2 MOF, 1 BOF

Recommended Starter Set (6 Concepts)

When the user selects Starter tier or does not specify, generate these 6 as a minimum viable test:

1. TOF x Static x Pain Point
2. TOF x Video x Desire
3. TOF x Carousel x Social Proof
4. MOF x Video x Social Proof
5. MOF x Static x Desire
6. BOF x Static x Pain Point (with offer)

Phase 3: Asset Generation (Optional)

If asset generation tools are available and the user requests assets:

Static Images

1. For each static concept, compose an image generation prompt based on the brief's hook, visual direction, and brand context
2. Generate using nano-banana generate_image.py:

   # Feed static (1080x1080 = 1:1)
   python3 scripts/generate_image.py "{hook visual description}, {brand aesthetic}, {product in context}" \
     --aspect-ratio 1:1 --output ./assets/ --filename tof-static-pain-v1.png
   
   # Stories/Reels static (1080x1920 = 9:16)
   python3 scripts/generate_image.py "{same concept adapted for vertical}" \
     --aspect-ratio 9:16 --output ./assets/ --filename tof-static-pain-stories-v1.png

3. Aim for smartphone-aesthetic quality — studio polish is not needed (smartphone beats studio 84% of the time in Stories)
4. Generate 1-2 variants per concept for A/B testing
5. Product cutouts on colored/gradient backgrounds: Use the 4-pass transparent PNG pipeline:

   # Generate product on white bg → remove bg → make transparent → composite
   python3 scripts/generate_image.py "{product}, on a simple white background" \
     --output ./assets/ --filename product-raw.png
   python3 scripts/edit_image.py \
     "Remove the background completely. Make the background fully transparent. Keep only the main subject with clean edges. Output as PNG with alpha transparency." \
     --images ./assets/product-raw.png --output ./assets/ --filename product-keyed.png
   python3 scripts/make_transparent.py ./assets/product-keyed.png \
     --output ./assets/ --filename product-transparent.png
   pngquant --quality=65-85 --force --output ./assets/product-transparent.png \
     ./assets/product-transparent.png

   Then composite the transparent product onto ad backgrounds using edit_image.py with the product + background as multi-image input.

Carousel Cards

1. For each carousel concept, generate individual card images using nano-banana generate_image.py with --aspect-ratio 1:1
2. Card 1 must hook with the problem or question — never lead with the product
3. Final card includes the CTA and offer
4. Maintain visual consistency across cards — pass all prior cards as reference images to edit_image.py:

   # Generate card 1, then use it as style reference for cards 2-5
   python3 scripts/edit_image.py \
     "Create card 2 of a carousel in the exact same visual style. {card 2 content description}" \
     --images ./assets/carousel-card-1.png --output ./assets/ --filename carousel-card-2.png

Video

1. For each video concept, generate a video using fal.ai if available
2. Focus the prompt on the first 3 seconds — this decides everything
3. Include text overlay direction in the prompt for hook text
4. If fal.ai video is not available, output a detailed script/storyboard instead:
   • Shot-by-shot breakdown
   • Timing (focus on 3-second hook)
   • Text overlay copy
   • Audio/music direction
   • Recommended length (15s, 30s, or 60s)

Phase 4: Output Summary

After generating all concepts, provide a summary table:

Follow with:

• Recommended test order — which concepts to launch first
• Performance benchmarks to watch — hook rate thresholds for video (kill below 25%, good 30-40%, scale above 40%), static conversion share targets
• Iteration guidance — when to kill, iterate, or scale each concept

Example Output

See examples/starter-set-example.md for a complete example of the 6-concept starter set generated for a premium butcher shop. Use this as a reference for the expected output format, level of detail, and quality bar.

Format Specs Quick Reference