Short answer: Start with a verified fact sheet and the product's one mechanism. Make one or two deliberate choices (an angle, a bias, a mindstate, a stage, a framework, a length or a processing route) and let everything else serve those choices. Then write one person, one problem and one turn, in plain speech, in the order Safety, Emotion, Logic, and close on something the viewer has earned.
Script Vibe is our internal script system: the frameworks, hook types and persuasion principles we use, pulled into one pipeline. This article walks through it and links to a deeper article on every part. The brands in the examples are made up.
What is the Script Vibe system?
Script Vibe is a five step pipeline for writing direct response ad scripts: build a fact sheet, make one or two picks, write, package the output, then check it silently before anyone films it.
The principle behind it fits in one sentence: do not sell the product. Clarify a pain the viewer already feels, connect it to a mechanism they can understand, prove it with evidence they can trust, and let the product be the obvious conclusion. Every rule below exists to protect that sentence.
| Step | What happens | Go deeper |
|---|---|---|
| 0. Fact sheet | Collect only what is true and provable | The mechanism |
| 1. Picks | Choose one or two levers, no more | See the table below |
| 2. Write | Hook, handoff, proof, earned close | Hook library |
| 3. Output | Package the script for whoever films it | Seven output modes |
| 4. Check | Score it privately, fix weak spots | This article |
Why do most direct response ad scripts fail?
Most scripts fail because they try to do everything at once.
The problem: A brief arrives with seven boxes ticked: social proof, urgency, the guarantee, aspirational, under 30 seconds, three benefits. The writer includes all of it, and the result reads like a spec sheet with a face. An earlier version of this system asked for every choice up front, and the scripts got worse for exactly this reason.
The solution: Make one or two choices on purpose and let the writer own the rest. A good ad is one person, one problem, one turn. Two picks are enough to name the turn. Everything else (the remaining biases, the stage, the framework, the length) gets chosen quietly to serve the picks and keep the story whole.
Step 0: What goes on the fact sheet?
A fact sheet is the list of things you are allowed to say. If a line in the script is not backed by the fact sheet, it does not go in.
It holds six things:
- Product, offer, category, price or promo. Nightfold is a weighted sleep mask at $38.
- The one core mechanism. The reason it works, stated plainly. For Nightfold: gentle even pressure plus total blackout that blocks the light leaks normal masks let through at the nose.
- Verified proof. Exact review counts, guarantees, named results. If you do not have them, you do not use them. In this article we write proof as placeholders, like "[your real review count]".
- Objections, pulled from reviews or surveys. "Will it feel heavy?" is a real objection. "Is it high quality?" is usually a guess.
- The person on camera. Their real world, voice and story.
- Platform and compliance rules. No medical claims, no income promises.
The mechanism deserves special attention because it is the one constant in every ad you will ever make for that product. It is the proof engine, never the idea. We explain why in the mechanism article.
Step 1: What are the picks, and why only one or two?
A pick is a deliberate creative lever. You choose one or two, and the system fills in the rest.
| Pick | The question it answers | Article |
|---|---|---|
| Angle | What one product truth is this ad built on? | Ad angles |
| Bias (hook, body or close) | Which mental shortcut does each layer lean on? | The bias stack |
| Mindstate and motivator | Is the buyer chasing a gain or avoiding a loss, and what need sits underneath? | Emotional motivators |
| Entry stage | How much does the viewer already know? | Entry stage and the full funnel one shot |
| Framework | How does the script argue? | The Jake frameworks |
| Length | How much time does the argument need? | Ad script length |
| Processing route | Is this a fast cue ad or a thinking ad? | Peripheral vs central route |
Output mode (script, shotlist, voiceover and so on) is packaging, not a pick. So is how many concepts you want.
Objection handling, audience frames ("for new parents") and execution styles ("street interview") look like picks but are decisions the writer makes while serving the picks. The mechanism is not a pick either, because it never changes. Pick an angle and a close bias and you have told the writer what the ad is about and how it ends. That is a story. Pick seven things and you have handed over a checklist.
What does the system decide on its own?
Everything you did not pick gets chosen to fit what you did pick. The defaults follow a simple logic.
- Entry stage comes from the idea and the traffic. Cold audiences who have never heard of you skew Unaware and Problem Aware. Retargeting audiences (people who already visited or engaged) skew Product Aware and Most Aware. Stage basics are covered in awareness stages.
- Framework follows stage. Unaware pairs with Jake 3.0, Problem Aware with Jake 1.1 or 2.0, Solution Aware with Jake 2.0, Product and Most Aware with Jake 1.0.
- Length defaults to a standard direct response cut of 20 to 35 seconds unless the idea needs more room.
- Route defaults to peripheral capture, then a handoff to central argument, then a close that works on both routes.
You never list these choices in the script. You write with them.
What psychology structure is in every script?
Four overlapping structures run underneath every script, whatever the picks. They are not optional and they are not printed. They are just how the script is built.
1. Safety, then Emotion, then Logic. First the viewer has to feel this person is credible and not a pitch. Then they have to feel something (fear, pride, longing, belonging). Only then will they listen to facts. We borrow the language of the old "triune brain" model for this order. Neuroscientists no longer treat that model as a literal map of the brain, but as a writing rule the order holds up: lead with facts and nobody is listening yet.
2. A four beat emotional arc. We use brain chemical names as shorthand for four beats: cortisol (the problem, a small jolt of stress), oxytocin (a relatable person you trust), dopamine (the tease of a mechanism that might fix it) and serotonin (proof and relief). It is a writing shorthand, not a lab claim. The point is the sequence: tension, connection, anticipation, resolution.
3. The Transformation Ladder. Every script climbs four rungs: feature, benefit, outcome, transformation. For Clearbrook: the filter is visible (feature), so you can see it working (benefit), so you stop wondering whether your water is actually clean (outcome), so you become the person who knows instead of hopes (transformation). Scripts that stop at features feel like a product page.
4. The processing route arc. Peripheral capture in the first 3 seconds, a handoff line by 5 to 8 seconds that makes it personal, central proof in the body, and a dual route close. We explain this in full in peripheral vs central route.
These four line up: the cortisol beat is the peripheral capture, the oxytocin beat is the handoff, and dopamine and serotonin are the central body.
What does a finished Script Vibe script look like?
Here is a full script. The picks were an angle (the Ritual Upgrade: it turns a daily moment into the best part of the day) and a close bias (Now Bias: feel it tonight, or here, tomorrow morning). Everything else was chosen to serve those two.
Concept: The Seven Minute Morning
(0s) "I used to drink my coffee standing over the sink."
(3s) "Kids yelling, keys missing, half a cup gone cold by the time I found my shoes."
(6s) "And I realized the one thing I actually look forward to every day, I wasn't even tasting."
(11s) "So I tried making cold brew myself. Jars in the fridge, filters everywhere, it came out bitter every time."
(17s) "Then a friend handed me Copperbean. It's a concentrate. They steep it slow, for a long time, then filter it twice."
(23s) "That's why it's smooth. No bitter edge. I pour some over ice, add milk, done."
(28s) "One bottle is my whole week."
(31s) "Now I sit down for it. Seven minutes. Nobody's allowed to talk to me."
(35s) "Tomorrow morning could taste like that."
What the system chose silently: a Problem Aware entry (she knows mornings are chaos, not that cold brew concentrate is the fix), Jake 1.1 in a promotion mindstate, a Pattern Interrupt hook (an oddly specific, slightly sad image), Labor Illusion in the body (the long steep and double filter, which is also the mechanism), and a peripheral to central route with the handoff at 6 seconds ("the one thing I actually look forward to... I wasn't even tasting"). The close calls back to the opening image instead of saying "link below."
Notice what is missing: no feature list, no slogan. The mechanism lives inside her reasoning.
What rules does every script line follow?
These rules apply to every script, every mode.
- The hook is line one, 5 to 12 words. One hook, chosen to serve the picks. Check it privately for clarity, a gap worth closing, originality, value and emotion. Our hook library has the types.
- One person, one problem, one turn. If you need a second problem, you need a second ad.
- Product messaging lives inside the speaker's reasoning. "That's why it's smooth" beats "Copperbean features a dual filtration process."
- Personal and specific. "Jars in the fridge, filters everywhere" beats "I tried other methods."
- The close is earned and in character. Land on the payoff, the proof or a callback. Never "link below," never a slogan.
- If you mention it, show it. Audio starts at the first frame. Something visual changes in the first second and a half.
- Non actors get a rant, not lines. Give them an anchor card (the one idea, the beats in order, the phrases that must survive) and let them talk. Taglines and stat blocks go to a text overlay or brand voiceover, never a creator.
The litmus test is simple: would you watch this for 30 seconds if you knew it was an ad?
Step 3: How do you package the script?
The script goes out clean. Nothing inside it but the words and, where the mode needs them, the visual beats. No labels like "[LOSS AVERSION]" or "[HANDOFF]" inside the script, because creators read those labels out loud or perform them stiffly.
After the script come two bullets at most: the idea, and what the hook is doing. Add a single flag line only if something failed, such as a claim that could not be verified.
The same idea can be packaged seven ways depending on who receives it: a plain script, a script with keyshots, a set of five benefit spread concepts, a full shotlist, a static brief, a voiceover only script, or a trigger event script mined from customer language. Each is explained in seven ways to package an ad script.
Step 4: How do you check a script before filming?
Score it privately on nine things, then fix anything weak. The nine: hook, emotional arc, mechanism clarity, proof, organic feel, voice, close, platform fit and handoff clarity.
Then run the hard checks. Is every claim on the fact sheet? Does the product reveal match the entry stage? Does the handoff land by 5 to 8 seconds? Does the close offer both a reason and a cue? Is it one story from hook to close? Any dashes or age references?
None of this scoring goes to the creator or the client. If a script is already live and not converting, look for the closed gate among four conditions: does the viewer feel the pain, believe a solution exists, trust you, and find buying easy? Rewriting the hook fixes none of the last three.
What do you do when a script wins?
Build out from it before writing anything new: a customer, problem, education and lifestyle version (see the four versions of every winning ad), then the opposite mindstate (see promotion vs prevention) when it tires.
What are the most common mistakes?
- Writing from imagination. A plausible sounding stat that is not on the fact sheet is a liability, not a hook.
- Making the mechanism the idea. If every ad is "here's how our filter works," every ad is the same ad.
- Too many picks. The script turns into a checklist.
- Facts before feelings. Logic first and nobody is listening.
- A close that is a slogan. The viewer was in a story and suddenly gets a billboard.
Key takeaways
- Start with a fact sheet. If it is not on the sheet, it is not in the script.
- The mechanism is in every ad, but it is never the idea.
- Pick one or two levers. Let the rest serve them.
- Every script runs Safety, Emotion, Logic, climbs from feature to transformation, and hands off from cue to argument by 5 to 8 seconds.
- Keep scripts clean of labels and check them privately before filming.
- Expand winners into versions before writing new ideas.
FAQ
Do I have to pick something every time? Yes, but only one or two things. The picks give the script its point of view. If you pick nothing, the script drifts toward the most generic version of the product.
Which pick should a beginner start with? The angle. It decides what the ad is about. Read ad angles first, then add a close bias from the bias stack once the angle feels solid.
Does this only work for video? No. The same thinking powers static briefs, voiceover scripts and primary text.
How long should the script be? Default to 20 to 35 seconds for a standard direct response ad, then go shorter or longer when the idea needs it. The full guide is in ad script length.
If you want to know whether your current scripts are doing the work they should, we offer a free Meta account audit. It includes a full account audit plus net new reach and cumulative reach reports, so you can see which creative is reaching new people and which is just recycling the same viewers. You can request one at our contact page.
Check what you learned
Tap an answer to see if you got it right.
1. What do you start every script with?
2. How many levers should you pick?
3. What do you do with winners?
Ready for the full test? Take the Ad Scripts exam and get certified. Want this run on your account? Get my free audit.
Module 6: Ad Scripts · article 1 of 21
Next: What Is a Product Mechanism in Ad Scripts? The One Constant in Every Ad
About the author
Jake Mehani, founder of Creative Analytics
Jake has spent 15 years running Meta and paid social ads for consumer brands. Meta published his RBX Active Advantage+ case study, he taught the DMAnc Meta and TikTok ads masterclass, and he spoke at Motion's 2024 Creative Strategy Summit. His senior team runs media buying, creative and measurement like an in house team.