Why Instagram is Losing Gen Z: We Asked 100 Users to Compare TikTok vs. Reels

2026-02-19 · Source: Surge AI Blog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics, Project & Product Management · Depth: Intermediate, long

Summary

Meta's Instagram Reels is struggling against TikTok, as evidenced by a study where 100 Gen Z users preferred TikTok 9.6 times more. The study, conducted by Surge AI, involved participants rating the first 10 videos on each app across dimensions like humor, interestingness, and diversity. Users consistently cited TikTok's superior recommendation algorithms (55%), content freshness and diversity (38%), and lightheartedness (32%) as key advantages. In contrast, Instagram Reels was criticized for generic, performative, recycled, and influencer-heavy content, often featuring old TikToks. This "metrics paradox" suggests that Meta's reliance on simple metrics like "video watch time" leads to optimization for low-quality, "cringe" content, rather than genuinely engaging or delightful experiences, highlighting the need for human-centric evaluation in algorithm design.

Key takeaway

For product managers overseeing content discovery platforms, your reliance on basic engagement metrics like "watch time" risks optimizing for low-quality, recycled content that alienates younger demographics. You should integrate human evaluation into your algorithm development process to identify and prioritize content that genuinely delights users, fostering a more authentic and engaging user experience that can compete with platforms like TikTok.

Key insights

Human evaluation is crucial for optimizing content algorithms beyond simple engagement metrics to capture genuine user delight.

Principles

• Optimize for human values, not just clicks.
• Generic content algorithms alienate Gen Z users.

Method

A human evaluation method involves asking raters to compare content from different platforms across multiple qualitative dimensions (e.g., humor, diversity, serendipity) to inform algorithm training.

In practice

• Implement human rating panels for content quality.
• Prioritize content freshness and diversity.
• Focus on genuine engagement over watch time.

Topics

• Human Evaluation
• Recommendation Algorithms
• Social Media Metrics
• Content Quality
• Short-Form Video

Best for: Product Manager, Data Scientist, AI Product Manager, CTO

Related on AIssential

• Moving Beyond Engagement: Optimizing Facebook's Algorithms for Human Values — Optimizing social media for human-rated values rather than engagement metrics can foster healthier online interactions.
• How TikTok is Evolving the Next Generation of Search — TikTok videos offer authentic, visual, and concise information preferred by younger users over traditional search results.
• My most controversial opinion.... — Audiences prefer human-generated content over AI-generated media, viewing the latter as low-effort and transient.
• FlowTime: Towards Continuous Generative Watch Time Prediction via Flow-based Personalized Priors — FlowTime leverages continuous generative regression and personalized priors to enhance watch time prediction accuracy and address data multimodality.
• Unethical brain rot: why are millions watching AI fruits have affairs on TikTok? — AI-generated "fruit dramas" exploit psychological vulnerabilities and platform algorithms to drive massive engagement despite problematic content.
• Facebook adds new AI features, animated profile photos, and backgrounds for text posts — Facebook is integrating AI-powered creative tools to enhance user expression and re-engage younger demographics.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Surge AI Blog.