Progressive Amplification Waves

TikTok's distribution system works like a series of expanding test chambers. Every new video starts small and only earns wider distribution by proving it can engage successive audiences. This is how a video with zero followers can reach millions -- or stall at 200 views.

The four waves

Wave 1: Follower Test (200-500 views)

When a creator uploads a video, it is first shown to a sample of their active followers (those who have logged in within 24-72 hours). This is the 2026 follower-first model.

Advancement criteria:

• High completion rate (~70%+ in 2026)
• Strong save and share velocity
• Positive engagement signals (comments, rewatches)

Failure mode: If the video doesn't hit these thresholds within the follower cohort, distribution stops. This is the infamous "200-view jail" -- the video technically exists but gets almost no algorithmic push.

Wave 2: Interest Expansion (1,000-50,000 views)

If the video passes Wave 1, it enters a broader distribution pool. The system identifies users who share behavioral overlap with the initial engaged cohort -- people who watch similar content, even if they don't follow the creator.

Advancement criteria:

• Performance benchmarked against average engagement for similar content in the niche
• Completion rate and save rate must remain strong as the audience broadens
• Share rate becomes increasingly important

Key dynamic: The audience is no longer people who already like the creator. It's people the algorithm predicts will like the content. This is where the Two-Tower Neural Network does its most important work -- matching the video's content embedding to user embeddings of people who don't know the creator.

Wave 3: Viral Push (50,000-500,000 views)

The video enters the primary For You Page at scale. At this stage, share rate and save rate become the dominant signals, overriding simple completion metrics.

Advancement criteria:

• Share rate must stay above niche-specific thresholds
• Save rate must demonstrate utility or resonance
• The video must perform across diverse audience clusters, not just one demographic

Why shares dominate at this stage: Completion rate proves the video is engaging. Share rate proves it's worth spreading. The algorithm interprets shares as a signal that the content has value beyond individual consumption -- it's content that creates social currency.

Wave 4: Explosive Distribution (500,000+ views)

The video has cleared all niche benchmarks and is pushed aggressively across broader demographic lines and geographic regions. At this stage, the content is being shown to users with increasingly diverse interests.

Sustained criteria:

• Engagement velocity must remain strong across diverse audience clusters
• The video must not have triggered negative feedback at scale (mass "not interested" signals)
• Performance must hold as the audience becomes less targeted

The wave model visualized

Wave 1: Followers (200-500)
    |
    | 70%+ completion, strong saves/shares
    v
Wave 2: Interest-matched non-followers (1K-50K)
    |
    | Niche-benchmark performance
    v
Wave 3: Primary FYP (50K-500K)
    |
    | Share rate + save rate dominant
    v
Wave 4: Global distribution (500K+)
    |
    | Sustained engagement across diverse clusters
    v
Millions of views

The math of distribution

Each wave has a specific engagement velocity threshold. The algorithm doesn't just measure total engagement -- it measures the rate at which engagement accumulates relative to impressions:

• Wave 1: Completion rate > 70%, save rate > 5%, share rate > 2%
• Wave 2: Niche-benchmarked performance (varies by category)
• Wave 3: Share rate > 3% and save rate > 4% at scale
• Wave 4: Sustained velocity across demographic boundaries

If engagement velocity drops below the threshold for a given wave, distribution slows or stops. The video doesn't disappear -- it just stops being actively pushed.

The 2026 follower-first shift

Prior to 2026, TikTok operated almost entirely on an interest graph. New content was pushed to small test pools based on predicted behavioral affinity, completely ignoring whether those users followed the creator.

The 2026 change: existing followers serve as the mandatory algorithmic gatekeepers for nearly all organic reach. For the first 24-72 hours, distribution is restricted primarily to the creator's follower base. Only after passing this test does the video earn the right to reach non-followers.

Why the shift?

1. Quality filter -- Followers are more likely to engage positively, reducing noise in the testing phase
2. Creator incentive -- Encourages creators to build genuine follower relationships, not just chase viral moments
3. Spam resistance -- Makes it harder for spam accounts to game the system with mass uploads

Implications

For creators

• Building a genuine follower base matters more than ever
• The first 24-72 hours after posting are critical for initial engagement
• Content must satisfy existing followers before reaching new audiences
• Understanding your niche's engagement benchmarks is essential

For the algorithm

• The wave model prevents bad content from reaching millions
• It ensures that distribution scales with demonstrated quality
• It creates a meritocratic system where content quality matters more than follower count
• But it also means that truly novel content may struggle if the initial test audience doesn't resonate

For users

• The feed is pre-filtered for quality before it reaches you
• You're unlikely to see low-engagement content on your For You Page
• The diversity injection in the Multi-Stage Ranking Pipeline prevents monotonous feeds

See also

• Cold Start Problem - How the distribution system handles new content
• Video Completion Rate as Currency - The primary signal that determines wave advancement
• TikTok Signal Hierarchy - The full signal weighting
• Multi-Stage Ranking Pipeline - Where wave-based distribution fits in the architecture
• Why TikTok's Algorithm Feels So Good - Why this creates a satisfying feed