| name | hooks |
| description | Use when generating hooks, headlines, titles, and scroll-stopping openers for content. Also use when analyzing viral posts, Reels, TikToks, YouTube Shorts, or successful social examples to extract reusable hook patterns and improve hook guidance. |
Headlines & Hooks Engine
Generate scroll-stopping titles, subtitles, and opening hooks for articles, posts, and videos.
The Concert Theory (Mental Model)
Source: @aidan.melograna (reel, credit @briarcochran for the idea).
Imagine you walk into a stadium with 25,000 people watching you. You have 5 seconds to say one thing to the crowd. If what you say doesn't create an emotional reaction, they all turn around and leave.
That's a hook. Not a sentence — a survival event.
Three jobs of every hook (in this order):
- Stop the scroll — pattern interrupt, tension, or open loop
- Make it obvious who it's for — the right viewer should feel called out within 5 seconds
- Create just enough curiosity to keep them listening — withhold the payoff
The one question to ask before posting anything:
"If I said this on a stage, would anybody lean in to listen more — or would they just pull out their phone and scroll?"
If the answer is "scroll," the hook is dead. Rewrite it.
Two rules that fall out of this:
- Don't open with context. Open with tension. Context is what you owe the reader after they're listening, not before.
- First sentence must make them feel seen or challenged. If it doesn't, you're already cooked. "Seen" = you named their secret. "Challenged" = you contradicted what they believe.
Attention is leverage. If nobody stays, nothing else in the post matters — not the insight, not the CTA, not the proof. Content isn't about talking more. It's about saying the right thing first.
LinkedIn Hook Formula
LinkedIn shows exactly 2 lines before the "…more" button. That cutoff is the only conversion that matters. If readers don't click "…more", nothing else counts.
Constraints:
- 2 lines max, each ≤55 characters
- Never about the author — always about the reader or a universal tension
- Creates an open loop: unanswered question, contradiction, or bold claim
- No emoji openers, no hashtags, no personal achievements
5 pattern-interrupt techniques:
| Technique | Formula | Example |
|---|
| Contradiction | Say something that sounds wrong | "The worst posts get the most followers." |
| Specific number + twist | Number + unexpected result | "I deleted 500 connections. Reach went up 40%." |
| Direct accusation | Call the reader out | "You're writing for your mom, not your market." |
| Stolen thought | Say what they secretly think | "You already know your posts are boring." |
| Absurd reframe | Make the mundane dramatic | "Your hook has 1.2 seconds. Most waste all of them." |
Process: Generate 10 variants (2 per technique). Don't pick the "safest" one — pick the one that made you wince slightly.
Working Examples (Before → After)
These show the transformation. The before is what a first draft usually looks like. The after is what actually gets clicked.
Topic: AI stock research
- ❌ Before: "How AI Can Help You Research Stocks Better"
- ✅ After: "The Stock Market Has a Cheat Code. Most People Just Don't Know It Exists."
- Why: Before describes the tool. After sells the outcome + implies insider knowledge.*
Topic: Bloom app feature
- ❌ Before: "Bloom Now Has an AI Portfolio Analyzer"
- ✅ After: "I Let an AI Critique My Portfolio for 30 Days. It Found Things My Broker Never Mentioned."
- Why: Personal experiment framing = relatable. Specific timeframe. Implies betrayal by trusted source.*
Topic: Investing with AI
- ❌ Before: "Why Retail Investors Should Use AI Tools"
- ✅ After: "Hedge Funds Have Had This Advantage for 10 Years. You Can Have It Now for Free."
- Why: Us vs. them + democratization. "10 years" makes the gap feel real. "For free" handles objection before it forms.*
Topic: Building Bloom (founder angle)
- ❌ Before: "Lessons From Building a Fintech App as a Solo Founder"
- ✅ After: "6 Months Building a Fintech App Alone Taught Me One Thing Most PMs Never Learn"
- Why: Specific time + specific person (solo) + promises one concrete lesson. Curiosity gap.*
LinkedIn hooks for the same Bloom topic:
Line 1 (≤55 chars): "I replaced my analyst with an AI."
Line 2 (≤55 chars): "It found what 3 years of manual research missed."
Pattern: Contradiction → Proof. Reader hits "…more" to find out what it found.
UGC Ad Hooks
Hook patterns optimized for AI-generated UGC ads (TikTok/Instagram). These are paid ad formats — uploaded directly to TikTok Ads Manager, not posted organically.
Core Rules for Paid Ad Hooks
- Never show the app/product in the first 3 seconds. Hook with emotion, curiosity, or a pattern interrupt first. The product reveal comes after the hook lands.
- Pattern-interrupt visuals: Use provocative or unexpected imagery in the first frame to stop the scroll (within platform ad policies). The goal is comment bait — visuals that make people react, argue, or tag someone.
- Result first, interface last. The reveal is the ad. Show the outcome, score, rating, scan, insight, or transformation before explaining the app.
- Trigger the desire, not the feature. The hook should touch a core desire or insecurity: beauty, health, wealth, status, intelligence, identity, fear, belonging. A feature-only hook is too weak for cold traffic.
- One winning creative beats endless strategy. Once a hook/visual combo works, make 10-12 variants with the same hook structure and different person, asset, stock, story, or reveal. Do not restart from scratch.
Source: @athcanft (May 2026, Apr 2026 X guide)
Truth-First Credibility Hooks
Source: Naval Ravikant, Sell the Truth, nav.al/sell, May 2026.
Cold audiences resist being sold to. A credibility hook earns attention by telling the truth the seller is not supposed to say, then only pitching after trust is created.
Strong forms:
- Steer-away hook:
Don't buy [thing] if [disqualifying condition].
- Bad-fit hook:
[Product] is wrong for you if [honest exclusion].
- Limitation hook:
[Product] cannot tell you [thing people wish it could]. It can tell you [truthful capability].
- Rejected-example hook:
I wanted [thing] to pass. It didn't.
- Truth-before-offer hook:
The uncomfortable truth about [desired outcome] is [constraint].
- Expert simplification hook:
Here's the simplest way to know if [choice] is actually good.
- Walk-away hook:
If this doesn't resonate in 10 seconds, skip it.
Why it works:
- It signals long-term thinking instead of quota-chasing.
- It filters for the person the message will actually resonate with.
- It borrows trust from the willingness to say no.
- It makes the product feel like the conclusion, not the agenda.
Bloom applications:
Don't use Bloom if you want someone to tell you what to buy.If your portfolio is just one S&P fund, you may not need this.The uncomfortable truth: most stock research is just confidence theater.Here's the simplest way to know if your stock thesis is real.
Rule: only use this if the exclusion is true. Fake honesty is just another sales tactic, and sharp audiences can smell it.
Customer-Language Hook Mining
Source: @adamtaylorl X article on beating Meta creative fatigue, May 2026.
The best cold hooks sound like they came from the customer's group chat, not the brand's positioning doc. Before writing hooks, mine reviews, Reddit threads, comments, post-purchase surveys, and viral organic posts for exact phrasing, failed solutions, and inside jokes.
Use the language in the hook when it proves you are one of them:
- exact complaint:
I keep buying dips and somehow still feel late
- failed solution:
I tried reading 10-Ks. I still missed the obvious risk.
- inside joke:
My portfolio is just vibes and NVDA at this point
- identity phrase:
for people who want to invest but hate finance bro advice
Rules:
- Keep the awkward specificity. Sanitizing customer language makes it generic.
- Use one real phrase per hook, not a pile of audience slang.
- Match the visual to the phrase. If the hook says
back on the golf course, the image cannot be a glowing knee joint. Congruency beats cleverness.
- Save source snippets with the hook so winners can be traced back to the language that created them.
Bloom applications:
My portfolio is up, so why does it feel like I got lucky?I bought the "safe" dividend stock. Bloom found the part I skipped.If your stock thesis is just "AI," run this check first.
Audible Stim Cold Open
Source: @Jibran_05 X post on UGC Reels audio hooks, May 2026.
For Instagram Reels UGC, treat the first 0.5 seconds of audio like a visual first frame. A tiny, natural sound can break scroll autopilot before the viewer has processed the words.
Working stim examples:
- ice clinking while sipping water
- phone slam or phone set-down
- pen click
- loud keyboard typing
- cap pop, bag crinkle, page flip, marker squeak, coin drop
Why it works:
- Immediate sensory interrupt: the sound starts faster than language comprehension.
- Native texture: it feels like real phone footage, not an ad bed.
- Micro-loop: the viewer asks what just happened and stays long enough for the spoken hook.
- Attention layering: visual hook plus audio hook gives the first second two jobs, not one.
Rules:
- Put the sound in the first 0.0-0.5 seconds, before or under the first word.
- Make it diegetic. The sound should plausibly come from the action on screen.
- Keep it short. This is a click, clack, sip, slam, or type burst, not a sound effect montage.
- Do not use random SFX that make the ad feel edited. The point is native sensory friction.
- Pair it with a real hook. Audio polish cannot save a weak premise.
Bloom applications:
- phone set-down →
I ran my portfolio through Bloom and it caught this
- keyboard typing burst →
I asked Bloom to find the problem in NVDA
- pen click over ticker list →
Three stocks people keep sending me. One failed the check.
- calculator tap →
Your dividend stock might be hiding this
Pre-Posting Hook Diagnostics
Source: @johnvirality on Higgsfield Virality Predictor / brain_activity, May 2026.
When evaluating finished short-form cuts before posting, separate four questions:
| Metric | What it diagnoses | Fix if weak |
|---|
| Audio interrupt | First 0.5 seconds break scroll autopilot | Add a native diegetic stim: pen click, phone set-down, ice clink, typing burst, page flip |
| Hook score | First 3 seconds stop the scroll | Stronger first frame, closer face, clearer text overlay, faster motion, sharper pain expression |
| Hold rate | Middle keeps attention after the hook | Remove dead air, add payoff beats, keep each line creating the next question |
| Viral potential / share signal | Viewer has a reason to send it | Add social currency: warning, identity, status, humor, "this is so me," or useful shortcut |
Use predictor scores as a triage tool, not taste. A video can score well and still feel like AI slop. Human taste gets veto power.
Variant rule: test 3-5 first-frame approaches on the same script before changing the whole concept: face-forward, environment-forward, direct eye contact, looking away, frustrated expression, vulnerable expression, text-first, action-first.
Bigger-Picture Preamble Hook
Source: Naval Ravikant, Sell the Truth, nav.al/sell, May 2026.
For sophisticated audiences, do not jump straight into the pitch. Start with the larger context, story, or problem situation, then make the small action feel obvious.
Structure:
larger truth → audience-specific implication → small action
Examples:
Retail investors don't lose because they're dumb. They lose because the tools hide the wrong things. Run one portfolio check.AI did not make stock picking easy. It made bad research easier to mass-produce. Check the thesis before you trust it.The market is not short on opinions. It is short on clean reasoning. Start there.
Use when selling to smart, skeptical, or high-intent audiences. Skip when the format needs a pure visual first-frame hook.
Investigation Arc UGC Hook
Source: @adamtaylorl X article on beating Meta creative fatigue, May 2026.
A strong anti-ad UGC structure is the Investigation Arc: creator reacts to a familiar expert, product, claim, or habit, discovers a better answer, then drops proof early enough to spike curiosity.
Structure:
react to familiar thing → find contradiction → investigate → early proof drop → product as tool/discovery
Why it works:
- It feels like discovery, not a pitch.
- The familiar target earns instant context.
- The contradiction creates the open loop.
- Early social proof gives the viewer a reason to trust the investigation before the CTA.
Bloom shells:
I checked [guru]'s stock pick because something felt offEveryone was calling [ticker] obvious. Bloom disagreed on one thing.I tried to prove this Reddit trade was smart. The data made it worse.I thought my broker screen showed everything. Then I ran the portfolio check.
Brief creators with idea plus freestyle: give the hook, proof points, and customer phrases, then let them riff in their own words. Over-scripted UGC loses the native texture that makes the investigation credible.
"this is ur reminder that..." format
Casual, direct, creates urgency without being pushy. 1.3M views on @simonecanciello's example. Best for AI talking-head UGC where the vibe is relaxed and authentic.
Working examples for Bloom:
- "this is ur reminder that you don't need a financial advisor"
- "this is ur reminder that checking your portfolio 10x a day isn't a strategy"
- "this is ur reminder that your savings account is losing to inflation"
- "this is ur reminder that you can research stocks without a Bloomberg terminal"
Template: this is ur reminder that [pain point everyone ignores] [and optional solution hint]
Before/After Transformation (visual-only hook)
No words in the first 3 seconds. The visual transformation IS the hook. High-performing TikTok ad format for app installs per @athcanft.
Format: "before" (3-5s) → cut → "after/reveal" (3-5s) → app screenshot CTA (2s). 12s total.
What counts as the reveal: face rating, body scan, food calorie breakdown, score, risk flag, AI insight card, portfolio diagnosis, or any result someone can understand in one frame.
Bloom example: confused investor staring at noisy ticker/chart → cut → clean Bloom AI insight/risk score/portfolio diagnosis → app screenshot CTA.
Hook shells:
I ran [thing they care about] through [analysis] and it caught this[Ticker/guru/portfolio] looked fine until the score came backThis is what your broker screen is hidingBefore Bloom: [confusion/noise]. After Bloom: [clear insight/reveal].
Production principle: Make rough, native-looking variants fast. Trending sound, clear reveal, one-line overlay, no polished ad sheen. The moment it feels like an ad instead of a TikTok, the hook gets weaker.
See also: VISUAL-HOOKS.md Sources 7 & 8 for full visual hook breakdowns.
Familiar Brand as Villain (Product Test Reveal)
Source: @oasishealthapp — 30M views across 232 Reels, $23K/month MRR, 100K+ followers. Solo founder, identical format every video.
The formula: take a brand everyone knows, reveal a hidden problem using data, position your product as the tool that exposed it.
Structure:
Familiar Brand (0-2s) → Hidden Problem with Data (2-8s) → Score/Rating Reveal (8-12s) → "Here's what I use instead" (12-15s)
Why it works:
- Familiar brand = instant recognition = scroll stop
- Data as drama: the score reveal IS the content
- Educational framing bypasses ad fatigue ("I tested X" reads as journalism, not promotion)
- The product is discovered as "the tool I used," never pitched directly
- Identical structure scales infinitely: same format, different brands, daily
Bloom applications:
- "I tested ChatGPT's stock picks" → bad performance data → "here's what real AI analysis looks like"
- "I ran Robinhood's top picks through Bloom's AI" → missed signals → AI confidence scores
- "I asked Jim Cramer's top pick to Bloom" → what the AI flagged that he didn't
- "WallStreetBets said buy [ticker]" → what the data actually shows
The villain shortlist (brands everyone has an opinion on):
ChatGPT, Robinhood, WallStreetBets, Jim Cramer, Mad Money, Motley Fool, Yahoo Finance, CNBC
Key principle: the villain must be something the audience currently trusts or uses. The cognitive dissonance between "I use this" and "it's actually bad" is what drives engagement.
Self-sustaining loop: Content generates app usage → usage generates data → data generates more content. Oasis runs lab tests that become videos. Bloom runs AI analysis that becomes videos.
After-Before Product Proof
For visual products, try the WayShot order: result first, ordinary original second, product mention last. The hook is a social observation or tool-confusion line, not a feature claim.
Best hook shells:
That friend who thinks everything is [aesthetic / cinematic / optimized]she [edits / tracks / analyzes] too much → reveal: damn you're rightWhat [camera / tool / analyst] did you use? → reveal: Oh it's my phone
Use this when the content can prove value through contrast. Load references/after-before-product-proof-sge.md for templates and adaptations. (Source: Social Growth Engineers WayShot case study, May 2026)
References
This skill content is modularized into reference docs for readability.
Output Format
Generate 10 headlines for the given topic or content. Present as a numbered list:
- ⭐ Best headline here — Rationale: why this one wins (hook strength, clarity, curiosity gap)
- Another headline — Rationale: one line on why it works
- Another headline — Rationale: ...
...
Rules:
- Star ⭐ the single recommended pick
- Each headline gets a 1-line rationale explaining the mechanism (curiosity gap, pattern interrupt, specificity, etc.)
- Mix formats: questions, statements, "How to", lists, contrarian takes
- After presenting, ask if the user wants variations on any specific headline or adaptation for a different platform
