how i find VIRAL faceless youtube video ideas with fable 5

the script idea is half the battle. and most faceless creators lose it before they write a single word.

they recycle the same 10 angles their competitors used last month. they browse trending topics and pick whatever has the highest search volume. they open an AI model and say "give me video ideas about [niche]" and take whatever surfaces.

then they wonder why their retention graphs look identical to every other channel in their space.

the research phase determines the ceiling for every metric that follows. the hook strength, the pacing structure, the curiosity loop placement, the outro conversion rate. every one of those is downstream of the idea you chose and the angle you entered from.

i've been running a specific ideation system using fable 5 for the last two months. it replaced a manual process that took 3+ hours and produced better results in 30 minutes. today i'm documenting the entire SOP exactly as i run it.

why fable 5 specifically for ideation

most creators use AI to generate ideas. i use fable 5 to evaluate ideas against retention data before i commit to writing a single word.

fable 5 is mythos-class intelligence. the longer and harder the task, the larger its lead over every other model. ideation isn't about generating a list of topics. ideation is about running complex analytical comparisons across multiple data sources to identify which specific angle, entry point, and structural approach will produce the highest retention for a given topic.

the three things fable 5 does here that sonnet 5 and opus 4.8 don't do as well:

• cross-referencing depth. i feed fable 5 outlier transcripts, competitor channel data, search intent analysis, and trend signals simultaneously. the model holds all of this in context and cross-references patterns across sources with a consistency that previous models couldn't maintain across a task this complex.
• self-validating output. when fable 5 identifies a promising angle, it checks the conclusion against the data before presenting it. if the evidence is weak, it flags the weakness rather than confidently presenting a fragile recommendation.
• structural recommendations. fable 5 doesn't give me a list of topics. it gives me a topic with a recommended hook type, an entry point, a pacing suggestion, and a structural format recommendation, all justified against the patterns it found in the outlier data.

the SOP: 5 phases, 30 minutes total

phase 1: outlier identification (5 minutes)

before i research any topic, i identify what's already working at an outlier level in the niche.

an outlier is a video that significantly outperformed its channel's average. a channel averaging 30K views that drops a 500K video has an outlier. a channel averaging 100K that drops a 95K video doesn't.

i connect nexlev MCP to the fable 5 project and run:

"search for channels in [niche] with 10,000-200,000 subscribers. for each channel, pull their last 30 videos. identify videos where the view count is at least 5x the channel average. return the top 15 outliers sorted by view-to-average multiple. include the video title, URL, view count, channel average, and publish date."

what i get back: a ranked list of the 15 videos that broke through hardest in the niche in the last 6 months. these are the structural blueprints worth studying.

phase 2: outlier transcript analysis (10 minutes)

i pull the transcripts of the top 5 outliers and feed all of them into fable 5 simultaneously.

"here are 5 transcripts from videos that outperformed their channel average by 5-20x in the [niche] space. do a deep structural analysis across all of them and identify the common patterns. for each transcript, analyse: hook architecture (psychological trigger used, where the first gap opens, whether it confirms or subverts the thumbnail expectation), pacing pattern (where it compresses, where it expands, where the shock moment lands), topic entry point (chronological, in media res, contrarian angle, modern reference), and the specific editorial decision that separates this video from a generic take on the same topic."

this analysis takes 8-10 minutes and produces something manual research never could: a cross-referenced structural blueprint of what's working right now.

not what worked 6 months ago. not what a guru said works. what the actual data from the actual niche shows is producing outsized results this week.

if you want to see how this phase feeds into the script brief that the writing phase builds from, i cover the full pipeline in the newsletter at fyreinteractive.co/newsletter.

phase 3: angle generation against the blueprint (5 minutes)

this is where most ideation systems stop at "topics." mine doesn't generate topics. it generates angles.

a topic is "the history of the cold war." an angle is "the 7 minutes that almost ended the world, and why the person who saved it was a 44-year-old soviet submarine officer who disobeyed a direct order."

the topic is generic. the angle is a script.

i take the structural patterns from phase 2 and feed them back into fable 5:

"based on the outlier patterns above, generate 15 script angles for [niche]. each angle must include: the specific entry point (where in the topic's timeline or narrative the script begins), the hook type that matches the outlier blueprint (contradiction, in media res, problem agitation, stakes escalation, or pattern interrupt), one sentence describing the curiosity gap the hook opens, and a note on which format the angle suits best (celebrity documentary, true crime, educational explainer, case study, countdown)."

15 angles. each with a structural foundation. each mapped against what's already proven to work in the niche.

phase 4: retention viability scoring (5 minutes)

not every angle is worth writing. some look strong on paper but have structural problems that would tank retention.

i run a viability check on the top 5 angles:

"for each of these 5 angles, evaluate: (1) can the hook confirm the viewer's expectation within 15 seconds while maintaining the curiosity gap? (2) does the topic have enough distinct sub-sections to sustain 5-6 mini payoff loops without the body going flat? (3) is there a natural grand payoff that the entire structure builds toward, or would the ending feel arbitrary? (4) does the topic have specific data, names, dates, or events that can serve as proof anchors throughout the body, or is the topic mostly abstract? score each angle 1-10 on retention viability and explain the weaknesses of any angle scoring below 7."

this check kills about 40% of angles before i write a word. the angles it kills are the ones that would have produced a script with a flat middle section, a weak grand payoff, or a body that runs out of specific material and defaults to abstract filler.

those are exactly the scripts that produce the payoff void and the retention cliff at minute 4. the viability check catches them at the ideation stage instead of the editing stage.

phase 5: script brief assembly (5 minutes)

the surviving angles (usually 2-3 from the original 15) get turned into full script briefs. the brief is not the script. the brief is the architectural document the script gets built from.

"for the highest-scoring angle, produce a complete script brief including: the hook (first 50 words, written out, using the recommended hook type), the entry point and why it works, 5-6 mini payoff sections in order with a one-sentence description of each payoff and a bridging sentence into the next, the grand payoff (what the viewer gets at the end that no competing video delivers), the pacing map (which sections compress, which expand, where the shock moment lands), and the viewing experience (two sentences on how the viewer should feel at 1 minute, 5 minutes, and the end)."

the brief takes 5 minutes to generate and produces a document that makes the scriptwriting phase 3x faster because every structural decision has already been made.

how the idea determines everything downstream

the reason i spend 30 minutes on ideation instead of 5 minutes browsing trending topics:

the angle determines the hook type. a case study angle needs a contradiction or in media res hook. an educational angle needs a problem agitation or stakes escalation hook. choosing the wrong hook type for the angle produces a mismatch that kills retention at second 20 regardless of how well the hook is written.

the entry point determines the pacing structure. an in media res entry (starting in the middle of the most dramatic moment) requires a flashback structure with compression at the start and expansion when the context fills in. a chronological entry requires escalating stakes because the narrative needs to build momentum from a low starting point.

the sub-section count determines whether the payoff void appears. if the topic only has 3 distinct sub-sections, a 10-minute script will have 2-minute flat zones between payoffs. the viability check catches this and either recommends a shorter format or eliminates the angle.

the grand payoff determines the outro. the strongest outros create a curiosity gap into the next video. if the grand payoff doesn't naturally connect to another topic on the channel, the outro will feel bolted on. the brief identifies this connection before the script is written.

every structural decision in the finished script traces back to a decision made during ideation. the creators who skip ideation and go straight to writing are building without a blueprint. the structure wobbles. the retention graph shows exactly where.

the full system, all 21 skill files including research, ideation, the scriptwriting architecture, the retention coaching, and the authenticity audit, is at fyreinteractive.co/facelessos.

(8,000+ scripts. $5M+ generated for clients. 200M+ views. 21 claude skill files trained on pattern data across 50+ niches.)

haris