
How Fandom Strengthens Community Safety and Unlocks Advertiser Growth with Coactive’s Multimodal AI
About Fandom
The World’s Largest Fan Platform
Fandom reaches nearly 350 million unique visitors per month across 250,000+ wikis. From Game of Thrones to My Little Pony, fans turn to Fandom for knowledge, connection, and community. The platform also powers advertising for leading brands across entertainment and gaming.
The Scale of the Challenge
Millions of Uploads, Thousands of Risks
Every month, Fandom users upload more than 2.4 million images and 125,000 videos. Even if only 0.1% were harmful or unsuitable, the volume created:
- Safety risks for users exposed to disturbing content
- Complexity for advertisers, who need confidence that their campaigns appear only next to appropriate material
- High operational cost, with 500+ contractor hours per week spent on manual moderation
Florent Blachot, VP of Data Science & Engineering at Fandom, put it simply:
Partnering with Coactive has transformed moderation from a bottleneck into a competitive advantage. We can now manage millions of uploads with confidence, protect our fans, and deliver the brand safety advertisers expect.
From Manual Burden to Automated Precision with Coactive
Coactive integrated directly into Fandom’s moderation stack, applying multimodal AI to images and video:
- Automated moderation: 100% of images and videos reviewed instantly, with only edge cases surfaced for human review
- Brand suitability filtering: Advertisers gain confidence their campaigns run only next to content aligned with their values
- Quality assurance: Blurry, low-quality, or irrelevant uploads are flagged for re-upload
- Enterprise-grade deployment: Coactive runs in a self-hosted environment, meeting Fandom’s security and scale requirements
Building Advertiser Confidence with Proactive Brand Suitability
Fandom worked with Coactive to understand gray area content that may be appropriate for fans but not for certain brands and surfaced video content that was not yet monetized to:
- Expand brand-safe inventory across images and video, ensuring advertisers never run adjacent to inappropriate content
- Maintain uninterrupted access to demand partners and DSPs (demand-side platforms advertisers use to buy inventory programmatically) by integrating Coactive into their brand safety and suitability framework
- Decide on contextually appropriate content to monetize
Timothy Quievryn, Director of Community Safety, explained:
Community safety is always our first responsibility. Coactive strengthens that foundation and, at the same time, reassures advertisers that their campaigns are appearing in trusted, brand-appropriate contexts.
Looking Ahead
Expanding Brand Suitability and Modalities
Fandom and Coactive are exploring several next steps to deepen impact including scanning the historical backlog of 500M+ images to improve safety and remonetize, moderate new types of content, and boost relevance and performance of contextual advertising.
Download this Case Study
Coactive AI helps data teams extract insights from unstructured image and video data. It integrates visual data with familiar SQL and big data tools, using pre-trained models for trend analysis, content moderation, search, and mapping.