The Viral DNA Method: 6 scores every content creator should know

The short answer

Viral DNA is a 6-score framework for measuring whether a TikTok creator's formula is portable to a different account. The scores are: viral DNA (composite), replicability, originality, consistency, hook reuse rate, and audience fatigue. Each is grounded in measurable signals — engagement variance, hook similarity (Jaccard), face dependence — not vibes. Hooklayer's analyze_account returns all six on every call with cited evidence.

Most existing creator-analysis tools return follower count and engagement rate. Both are useless for the question that actually matters when you're planning content: "Can I copy what this creator is doing, or will it fail without their face?"

Why "viral" alone is the wrong metric

A 5M-view video tells you the format worked for that creator. It doesn't tell you whether the format will work for you. Tori Dunlap's "I'll have $6M by retirement and I'm only 26" hook pulled 7.8M views. Copy it as a 21-year-old with no credentials? You'll get clocked as cosplay within 100 views.

The non-portable formula problem is the hidden cause of failed content pivots: a creator copies a viral structure that depends on traits they don't have (credibility, reach, face brand, niche authority), and the algorithm punishes the inauthenticity.

The Viral DNA Method exists to separate the portable from the non-portable before you record a single video.

The 6 scores

1. Viral DNA Score (0-100, composite)

The headline number. Rolls up consistency, originality, and momentum. Calibrated against named anchors:

• 90+ = elite (consistently viral, e.g. MrBeast, Humphrey Yang)
• 70-89 = strong (reliably solid views, e.g. niche creators with 1M+ followers and steady output)
• 50-69 = mid (good days and bad days)
• <50 = inconsistent

What it doesn't tell you: whether you can copy them. That's what the other 5 scores are for.

2. Replicability Score (0-100)

The most important score in the framework. Answers: how face-dependent vs format-dependent is this creator's formula?

• 80+ = format-dependent. Anyone can copy. Example: A "Mid-Action Drop" hook with a $40K vs $1K compound interest reveal works in any voice.
• 40-79 = mixed. Format travels, persona doesn't. Most creators land here.
• <40 = face-dependent. Won't survive without the creator's specific identity. Talking-head pundits and credential-flex creators usually score here.

Cited signals (from Hooklayer's analyze_account response):

• face_dependence (0-10) — talking-head ratio across top 5 videos
• format_dependence (0-10) — structural beats present without persona

3. Originality Score (0-100)

Are they originating hooks, or recycling everyone else's?

• 80+ = template originator. First to invent the formula others copy.
• 50-79 = format master. Doesn't invent, but uses common patterns better than anyone.
• <50 = derivative. Copies what's already working.

This isn't a judgment — derivative creators can still go viral. But originality matters when you're deciding who to learn from. An originator's formula is more durable; a derivative's may be one trend cycle away from collapse.

4. Consistency Score (0-100, deterministic)

Computed from engagement-score variance across the top 5 videos. Not AI-generated — pure math. Lower variance = higher consistency.

• 80+ = highly predictable performance. Their formula is dialed in.
• 50-79 = solid but with outliers.
• <50 = boom-or-bust. One viral hit, four flops.

A high consistency score is the strongest indicator a formula is replicable, because it means the creator has tested it enough to remove the variance. If their consistency is 90, their structure is doing the work — not luck.

5. Hook Reuse Rate (0-1, deterministic)

How often does this creator recycle the same hook formula across their top 5? Computed via Jaccard similarity across opening lines.

• 0.0-0.2 = unique openers. Every hook is fresh.
• 0.2-0.5 = pattern-aware. Reuses a structure but varies the surface.
• 0.5+ = repetitive. Same hook, different topic.

Counterintuitively, a moderate hook reuse rate (0.2-0.4) is usually the strongest signal. It means the creator has a recognizable structure that travels — that's the thing you want to copy.

6. Audience Fatigue (stable / rising / declining)

Are recent videos under- or over-performing their average?

• rising = recent videos crushing the average. The formula is heating up.
• stable = consistent performance. Safe to copy now.
• declining = the formula is decaying. Whatever made it work is wearing out. Don't bet on it.

How to read the 6 scores together

The pattern that matters: high replicability + high consistency + stable audience fatigue + moderate hook reuse rate.

That combination tells you: their formula works, it's structural (not face-dependent), it's repeatable, and the audience hasn't gotten tired of it. That's the creator to learn from.

The "non-portable formula" warning

Every Hooklayer analyze_account response includes a would_fail_because field. This is the explicit warning: what specifically would break if you copied this formula without the creator's specific traits.

Examples from real responses:

• "If you copy Tori's age-flex hook ('I'm a millionaire and I'm only 26'), the algorithm reads it as cosplay because the credibility signal doesn't match the content depth."
• "If you copy Humphrey's dialogue-scaffolding without his calm-authority delivery, the structural beats land but the voice doesn't — viewers feel something is off without being able to name it."
• "If you copy MrBeast's high-budget production, the format collapses to a smaller scale; you can't \"do\" his videos with a smaller budget."

The warning IS the value. A "you should copy this" recommendation is useless without "and here's the trap."

How to apply this in practice

1. Pick 3-5 creators in your niche. Don't aspire to MrBeast — look at the people one tier above you.
2. Run analyze_account on each. That's 15-25 credits total. Free tier covers it.
3. Sort by replicability_score. The highest-replicability creator is who you learn from.
4. Read their steal_map. Hooklayer extracts 3 transferable patterns with concrete how-to lines and example_remix per pattern.
5. Read the would_fail_because. That tells you what NOT to copy.
6. Use viral_remix on their highest-replicability video. The script generator preserves their structural DNA in your topic.

What this method is NOT

• Not a content-quality predictor. A high viral DNA score for the source creator doesn't mean your output will go viral. It means the formula is copyable — you still need to execute.
• Not a substitute for taste. Scores tell you which formulas are portable. They don't tell you which ones FIT your account. That's a judgment call.
• Not exhaustive. There are content categories (deep storytelling, original investigations, slow-burn series) where these metrics don't apply. The method is calibrated for short-form viral content, not long-form auteur work.

The Viral DNA Method is built into every Hooklayer analyze_account response. Try it on the playground (no signup), or read the tool reference for the full schema.