Viral Content Detection
Share
A technology-driven content system that automatically detects and repackages viral content before it hits mainstream platforms, developed by Emerson Sparks for Dose Media.
Core System Components:
-
Content Detection:
- Monitors Reddit, Imgur, and 4chan to find trending content early
- Uses engineering team to build automated detection systems
- Identifies content before it reaches Instagram, Facebook, or Twitter
-
Content Production Pipeline:
- 18 engineers and 4 writers to process content
- Writers create summaries of detected trending content
- Automated AB testing system for headlines and framing
-
Testing & Optimization:
- Creates multiple variations of headlines and content frames
- Tests content on small audiences (5,000-10,000 people)
- Automated system identifies winning variations
-
Distribution:
- Pushes winning content variations to audience
- Focuses on rapid deployment to maximize virality
- Uses data feedback to inform future content strategy
-
Team Structure:
- Heavy emphasis on engineering (18 engineers)
- Lean content team (4 writers)
- Technology-first approach to viral content
The system successfully generated high traffic but was vulnerable to platform changes, particularly Facebook algorithm updates that deprioritized viral content.
45:24 - 46:50
Full video: 53:51SP
Shaan Puri
Host of MFM
Shaan Puri is the Chairman and Co-Founder of The Milk Road. He previously worked at Twitch as a Senior Director of Product, Mobile Gaming, and Emerging Markets. He also attended Duke University.