Solution · Ad Creative Audit
Ad Creative Duplicate Detection and Creative Audit
MediaLayer detects duplicate, reused, and modified ad creatives across campaigns, platforms, and media libraries — using fingerprint-based matching that survives re-encoding, watermarks, and minor edits.
The problem
Where this hurts in production
Duplicate creatives across campaigns
The same image or video runs under different campaign IDs, advertiser accounts, and platforms. Pixel hashes miss re-encoded copies. Manual review can't keep up with volume.
Modified ad variants
Old creatives resurface with swapped logos, cropped frames, or new audio overlays. The metadata changes — the content doesn't. Fingerprint matching catches what filenames miss.
Cross-platform creative overlap
The same asset runs across display, social, and video platforms under different IDs. Without content-based matching, cross-channel creative governance is guesswork.
Creative governance at scale
At 50K+ creatives per quarter, manual creative QA is impossible. Automated fingerprint matching surfaces reuse, variants, and duplicates before they hit brand safety or billing issues.
How MediaLayer fits
How MediaLayer fits this workflow
MediaLayer uses dHash fingerprinting and LSH band indexing to match ad creatives at scale. Two images with the same visual content — regardless of re-encoding, minor cropping, or watermark overlays — produce near-identical fingerprints. A similarity search against your indexed creative library surfaces all matches in milliseconds.
For video creatives, frame-level fingerprinting detects duplicate segments even when only part of a video is reused. For audio, spectrogram peak hashing identifies the same audio track across transcodes, remixes, and format changes. One consistent API handles all three media types.
The workflow is straightforward: index your approved creative library, then submit incoming creatives as search queries. Get back ranked matches with similarity scores. Build governance rules, flag reuse above a threshold, or route matches to a human review queue.
Workflow example
From media in to match decision out
- 1
Index approved creatives
Submit your reference creative library via URL or CSV bulk upload. MediaLayer fingerprints each asset and adds it to your searchable index.
- 2
Submit incoming creatives
As new creatives arrive, submit them as search queries via POST /v1/search/{media_type}/url.
- 3
Retrieve matches
Get back ranked matches with similarity scores, confidence labels, and matched segments for video and audio.
- 4
Apply governance rules
Set similarity thresholds per use case. Flag exact duplicates, near-duplicates, or partial reuse for review or auto-rejection.
# Request
POST /v1/search/image/url
X-API-Key: your-api-key
{
"url": "https://cdn.example.com/new-creative-q3.jpg",
"top_k": 5
}
# Response
{
"query_media_id": "a1b2c3d4-...",
"results": [
{
"media_id": "f9e8d7c6-...",
"name": "summer-campaign-hero.jpg",
"similarity_score": 0.97,
"confidence": "high"
},
{
"media_id": "b3c4d5e6-...",
"name": "q2-display-banner-v2.jpg",
"similarity_score": 0.84,
"confidence": "medium"
}
]
}Real-world examples
Patterns we see in this space
Creative deduplication across campaigns
Detect when the same creative runs under different campaign IDs, buyers, or dates.
Modified variant detection
Flag creatives that are watermarked, cropped, brightness-adjusted, or rebranded versions of existing assets.
Cross-platform overlap audit
Surface creatives running across display, social, and video platforms that share visual or audio content.
Pre-flight creative QA
Check incoming creatives against an approved library before they enter the campaign workflow.
Ready to ship?
Try Match API or discuss scale.
Try Match API on RapidAPI, or talk to MediaLayer AI Labs about high-volume access, private deployment, and custom rate limits.
Public API access is distributed through RapidAPI. Enterprise direct API access is available only after onboarding.
Looking for something else? Contact us.