Meta AI Boosts User Engagement and Revenue
Meta CEO Mark Zuckerberg says AI-driven recommendation systems fueled a 5% rise in Facebook usage and 6% on Instagram in Q2. Smarter algorithms helped attract 3.4 billion daily users, lifted video watch time by 20%, and grew family of apps revenue to $47.1 billion. Threads also saw engagement gains from LLM integration.
Meta AI Boosts Engagement and Revenue
On its Q2 earnings call, Meta CEO Mark Zuckerberg credited advances in artificial intelligence with driving higher user engagement across the company’s apps. While some users call out fatigue from lower quality “AI slop,” Meta says smarter recommendation models are surfacing more relevant and original content.
AI-Powered Recommendations Drive User Time
Zuckerberg told investors that improvements to Meta’s recommendation systems led to a 5% increase in Facebook time and a 6% lift on Instagram within the quarter. By leveraging data signals like viewing duration, social interactions, and fresh content ranking, AI models are matching people with posts and videos they’re more likely to enjoy.
Strong Growth in Reach and Revenue
- 5% increase in time spent on Facebook this quarter
- 6% increase in time spent on Instagram this quarter
- 3.4 billion daily users across Meta’s family of apps, up 6% year-over-year
- Total family of apps revenue reached $47.1 billion, up 22% year-over-year
- 20% rise in video watch time after ranking optimizations and original content boosts
- Threads engagement grew thanks to the incorporation of large language models
Balancing Innovation with Quality
User complaints about “AI slop” highlight the risks of flooding feeds with generic or low-value content. Meta responds by refining training data sets, tightening quality filters, and rewarding original creators. This balance keeps the platform vibrant and trust intact.
Lessons for Tech Leaders
Meta’s results underscore the need for robust analytics, continuous model evaluation, and feedback loops. Organizations can drive measurable engagement gains by treating AI as a dynamic system that learns from real-world usage and constantly adapts to user preferences.
- Collect comprehensive user interaction data to power smarter algorithms
- Run iterative A/B tests to fine-tune recommendation parameters
- Implement quality gates to block low-value AI-generated content
- Monitor key metrics like time spent, content diversity, and creator satisfaction
Keep Reading
View AllMicrosoft Tests Copilot Smart Mode Ahead of GPT-5 Launch
Microsoft is internally testing a new Copilot smart mode that dynamically switches AI models for tasks, hinting at GPT-5 integration and unified intelligence.
Google Rolls Out AI Age Verification to Curb Underage Access
Google will use AI to detect US users under 18 and automatically apply content, ad, and app restrictions to enhance safety and compliance.
Meta Charts Path to Personal Superintelligence
Mark Zuckerberg outlines Meta's shift towards closed models for personal superintelligence while balancing open-source benefits and safety.
AI Tools Built for Agencies That Move Fast.
Want to boost engagement like Meta? QuarkyByte’s analytics approach can sharpen your recommendation engines, optimize content quality filters, and deliver measurable lifts in time spent and revenue. Dive into custom AI strategies designed for scalable growth.