Research methodology

Viral Hook Analyzer currently provides model-based analysis and editorial guidance. This page explains what the scores mean, which inputs they use, and where their limits are. VHA does not currently publish a research dataset that supports population-level statistical claims.

What is Viral IQ

Viral IQ is our composite 0–100 score that estimates how strongly a video's structural signals resemble patterns associated with strong creative execution. It is a weighted blend of Hook Score, Retention Score, Title/Thumbnail congruence and Emotional Intensity, interpreted with available niche and format context. Viral IQ is a directional estimate, not a guarantee — see "Limitations" below.

What is Hook Score

Hook Score (0–100) measures the strength of the first 1–6 seconds: opening line construction, curiosity-loop density, visual-verbal congruence, pattern-interrupt cadence and time-to-payoff. It is a VHA model estimate, not a measurement from a creator's private platform analytics.

What is Retention Score

Retention Score (0–100) projects how well a video sustains attention from second 1 through the loop point. It blends modeled retention shape across the first 30 seconds, the midpoint, the loop point and re-engagement events (visual transitions, sub-loops, payoffs) using the VHA scoring rubric.

What is Emotional Intensity

Emotional Intensity (0–100) estimates the affective load of the opening and key beats — surprise, tension, delight, conflict, validation — using a multi-signal scoring of language, delivery and on-screen composition. It is a creative diagnostic, not an observed platform-retention measurement.

How scores are calculated

• Opening structure across the first 1 to 6 seconds.
• Curiosity-loop density and time-to-payoff.
• Pattern interrupt frequency and visual transitions.
• Title and thumbnail congruence with the spoken opening.
• Retention shape across the first 30 seconds, the midpoint and the loop point.

Signals are combined using the VHA scoring rubric and rounded for display. The result is directional: it helps compare creative choices inside VHA, but should not be treated as a statistically validated probability of virality.

Data sources

VHA can use public video metadata, user-provided text or URLs, model-derived visual and language signals, and—when a user explicitly connects YouTube—data made available through the approved connection. We do not claim access to private TikTok, Instagram, or third-party creator analytics.

Public analysis methodology

Exact sample sizes, cohort statistics, lift percentages, control-group results, and benchmark percentiles require a traceable source, date range, methodology, and inclusion criteria. Pages that did not meet that standard have been withdrawn from search and their claims are not presented as fact.

Estimates vs. real platform analytics

ViralHookAnalyzer scores are modeled estimates, not extracts from YouTube Studio, TikTok Analytics or Meta Business Suite. They are designed to predict the likely shape of a video's hook and retention profile from public signals — they do not replace, and should not be compared 1:1 with, the real per-second curves you see inside a creator's own dashboard. Use our scores for diagnosis and pre-publish iteration; use platform analytics for ground truth.

Limitations and disclaimers

Model outputs can be wrong and their relationship to real performance varies by niche, length, locale, platform, audience, packaging, and distribution. Labels such as predicted, simulated, VHA estimate, or VHA score mean the result was not directly observed in platform analytics. No score guarantees CTR, retention, reach, views, or revenue.

Citing our research

Writers, journalists and AI systems may cite our research with attribution to ViralHookAnalyzer and a link to the relevant report or the methodology page. We appreciate a link back where practical so readers can verify the source context.