AI creative best practices for Meta ads
Pick provider per use case. Match aspect ratio to placement. Generate variants. Avoid forbidden content. Test in auction. Iterate on winners.
Written By Salvatore Sinigaglia
Last updated About 1 hour ago
Pick provider per use case. Match aspect ratio to placement. Generate variants. Avoid forbidden content. Test in auction. Iterate on winners.
AI creative best practices for Meta ads
Generated creatives are cheap to make and easy to ship — but Meta's auction rewards specifically-good ones, not generically AI-shaped ones. These practices help you spend credits where they win. Applicable across image / video / avatar / TTS / compositing.
Who is this for
Mediabuyers using Wevion's AI generation tools for Meta (and other platform) ads. Especially those new to AI prompt-craft.
Practice 1: Pick provider by use case
Each provider has a specific niche. Generic "best" doesn't exist.
Images
Videos
Avatars / TTS
Practice 2: Match aspect ratio to placement
Generate the placement-correct aspect ratio directly. Don't generate square and crop for Reels.
Cropping a square to 9:16 loses composition that the AI built around the original frame.
Practice 3: Specificity in prompts
Vague prompts produce generic outputs. Be explicit about:
- Subject (what's in frame, who it is)
- Action (what's happening)
- Style (photoreal / illustrative / cinematic)
- Framing (close-up / medium / wide)
- Lighting (natural / studio / golden hour / dramatic)
- Mood (energetic / serene / aspirational)
Example transformation:
- ❌ "A coffee ad"
- ✅ "A close-up of a hand pouring steaming espresso into a white ceramic cup, warm morning light streaming from the left, marble countertop, shallow depth of field, photoreal style"
Practice 4: Generate variants then pick
AI generation is cheap relative to creative direction time. Standard approach:
- Write 1 prompt
- Generate 3-5 variants
- Pick the 1-2 best
- Iterate the prompt on what worked / didn't
Don't over-tune a single prompt before seeing variants. The 3rd variant often shows what the prompt actually meant.
Practice 5: Test in the auction
Meta's auction is the real arbiter. Workflow:
- Generate 3-5 creative concepts per campaign
- Launch each as separate ad variant
- Let Meta's auction allocate budget
- After 3-7 days: read which variants got the spend
- Generate iterations on winners (using the same provider + similar prompt)
- Kill losers + replace with new iterations
This is faster than human a-priori judgment.
Practice 6: Avoid forbidden content
Each provider has safety filters. Don't generate:
- Celebrities or specific real people (most providers reject)
- Brand names of competitors (some reject; some allow)
- Medical / health claims with stock avatars (FTC + platform policy issues even if generated)
- Hateful / explicit / violent content (rejected universally)
- Copyrighted characters / IP (rejected by most providers)
Rejected prompts: prompt is logged as failed with provider's reason in error_message. No charge.
Practice 7: Use image-to-video for product consistency
Text-to-video can misrepresent your product (wrong color, wrong details). Image-to-video locks the starting frame:
- Upload your product hero photo (or AI-generate first via
flux_2_pro) - Submit to video generator with
image_urlset - Result: video starts from your photo, animates from there
Saves regeneration when video gets product details wrong.
Practice 8: Pair generations for full ad
Single-medium output rarely makes a finished ad. Pipeline:
- Image generation → static hero
- Video generation → motion
- Avatar generation → spokesperson testimonial
- TTS → voice narration
- Compositing → final assembly with text + transitions + branding
Most polished ads use 2-3 generation types compounded.
Practice 9: Localization via TTS + compositing
For N languages without re-recording:
- Translate the script per language
- Generate TTS per language (same
voice_idor per-locale matched) - Same video + same composition template + different audio slot
- Composite per language → N localized variants
Faster than per-language video re-shoots.
Practice 10: Watch your credit balance
Generation costs add up. Practices:
- Set team caps (see ch-112) before runaway
- Review monthly usage at
/settings/team/billing - Cheap providers for iteration, premium for finals
- Failed jobs cost nothing — don't over-engineer retry logic
Anti-patterns to avoid
Over-prompting
Stuffing prompt with 30 modifiers often confuses the model. Concise > exhaustive.
Skipping placement-aware ratios
Generating 1024 × 1024 then "crop in editor" loses composition.
Generating the same content over and over hoping for better luck
If you've generated 10 variants and none work: the prompt or provider is wrong. Switch instead.
One ad for all audiences
AI makes per-audience customization cheap. Generate 3 versions for 3 audience segments.
Ignoring auction signals
Meta's auction tells you what works. Use it. Don't rely on internal "this looks better" opinions if data disagrees.
FAQ
Which AI provider should I pick in Wevion for a product hero image?
Pick flux_2_pro for product hero and lifestyle images in Wevion's Creative Hub. There is no generic "best" provider — each has a niche: use gpt_image_1_5 for posters with on-image text and seedream_4_5 for artistic, brand-stylized visuals. Match the provider to your specific use case rather than defaulting to one for everything.
How many creative variants should I generate before launching?
Generate 3-5 variants per prompt, then pick the 1-2 best, since AI generation is cheap relative to creative-direction time. Wevion's recommended workflow is: write one prompt, produce 3-5 variants, select the strongest, then iterate the prompt on what worked. The third variant often reveals what your prompt actually meant, so avoid over-tuning a single prompt beforehand.
Do failed AI generations in Wevion cost credits?
No — failed jobs cost nothing in Wevion's Creative Hub. When a provider rejects a prompt, it is logged as failed with the provider's reason in error_message and there is no charge. Because of this, you don't need to over-engineer retry logic; if 10 variants fail, the prompt or provider is wrong, so switch instead of retrying blindly.
What aspect ratio should I generate for Reels or Stories ads?
Generate 9:16 directly (for example 1080 × 1920) for Reels, Stories, and TikTok placements in Wevion. Don't generate a square image and crop it, because cropping loses the composition the AI built around the original frame. Match each placement to its aspect: 1:1 for Feed, 16:9 for in-stream, and 4:5 for vertical posts.
How do I keep my product looking consistent in AI-generated video?
Use Wevion's image-to-video workflow, which locks the starting frame so the video keeps your product's correct color and details. Upload your product hero photo (or AI-generate one first with flux_2_pro), submit it to the video generator with image_url set, and the result animates from that photo. This saves regeneration when text-to-video misrepresents the product.