Ad Script Writing: A Four-Block Framework (With Template + Examples)

There is a popular belief that good ad scripts come from inspiration: the clever line, the unexpected angle, the moment a writer "gets it." That belief quietly ruins a lot of campaigns, because it treats a repeatable craft as a lottery. Ad script writing is closer to carpentry than to poetry. The best scripts are assembled from a small number of parts in a predictable order, and the parts are boring on purpose. Once you can name the parts, you stop waiting for a good idea and start producing testable ones on demand.

This guide lays out a four-block structure for writing ad scripts, gives you word-count budgets for each block, includes a fill-in template you can reuse, and walks through a teardown of one ad you have probably seen. None of it requires you to be funnier or more original than the next person. It requires you to put the right block in the right place.

The myth that the first line has to be brilliant

Writers obsess over the opening line because they have heard that attention is won or lost immediately. The part they misread is that a strong opener is not a clever opener. It is a recognizable one. The viewer is scrolling through a feed they chose, and your ad is the interruption. The only thing the first line has to accomplish is to make the viewer think the ad is about their situation, not yours. That is a much lower and more achievable bar than "be brilliant," and it is the difference between a hook that performs and one that just sounds good in the doc. If your openers keep falling flat, a library of tested hook patterns is a faster fix than waiting for a better idea.

Two failure modes show up over and over. The first is the brand-first open: "At Acme, we believe great teams deserve great tools." Nobody believes anything about your brand yet, so the sentence is spent on credibility you have not earned. The second is the feature dump, where the script lists what the product does before the viewer has any reason to want it. Features are answers. If you have not raised the question, the answers land on nobody.

The four blocks of an ad script

An ad script that holds attention does four jobs, almost always in this order. Each block exists to earn the right to the next one, which is why the sequence is not interchangeable.

1. Problem. Name the situation the viewer is already living in, precisely enough that they recognize themselves.
2. Agitate. Make the cost of that problem concrete, so it feels worth solving rather than tolerating.
3. Mechanism. Explain the one specific reason your thing produces a different result. This is where the product finally appears, framed as a "how," not a "what."
4. CTA. Ask for one action, in plain words, with a reason to take it now.

You may recognize the first, second, and fourth blocks as PAS, the problem-agitate-solution pattern that has been in copywriting books for decades. The framework here is not PAS with a new coat of paint. The change that matters is splitting "solution" into two separate jobs: the Mechanism, which explains why the thing works, and the CTA, which asks for the action. Most weak scripts collapse those into one vague gesture toward the product, which is exactly the seam where belief breaks. This same four-job spine sits under most of the UGC ad formulas the platforms keep surfacing, even when the delivery looks unscripted.

Word-count budgets that keep each block honest

The reason ad scripts run long is that writers give every block as much room as it asks for. A read-aloud budget fixes that. These are spoken-word counts, measured by reading the line at a natural pace, not by character count on the page. They are the calibration most short-form ads land in, and they keep the script tight enough to fit the attention you actually have.

• Problem: 12 words or fewer. If the opening situation takes longer than one breath to say, it is not specific, it is a paragraph. Cut until it sounds like something the viewer would say to a friend.
• Agitate: 10 to 18 words. One concrete consequence. The temptation is to stack three; resist it, because a single sharp cost lands harder than a vague pile.
• Mechanism: 18 to 25 words. This block gets the most room because it is doing the hardest work: making a claim believable. It is the only place where a slightly longer line is justified.
• CTA: 8 to 12 words. One action plus one reason it is safe or timely. Anything longer is usually a second ask sneaking in.

Added up, that is roughly a 15 to 30 second read for most placements, which is where short-form ads tend to perform. The budgets are a starting calibration, and you will adjust them for a slower B2B audience or a faster TikTok one. But starting with a budget you have to cut toward produces sharper scripts than writing freely and trimming later.

Block 1: Problem

State the viewer's situation in their words. The test is whether the viewer could have said this exact sentence to a colleague last week. If it sounds like marketing, rewrite it as a complaint. Compare "Managing social media is hard" with "You have nine tabs open and still don't know what to post." The first is a category. The second is a Tuesday.

Block 2: Agitate

Make the cost specific. Time, money, or the recurring low-grade dread of a task that never gets easier. The aim is recognition, not anxiety. You are confirming the problem is real and expensive enough to act on, then immediately handing over relief in the next block. Agitate that lingers turns into a downer; agitate that names one concrete cost sets up the mechanism cleanly.

Block 3: Mechanism

This is the block writers get wrong most often, because it feels like the place to list everything the product can do. It is the opposite. Pick the one thing that makes the result possible and explain it. "It writes your captions" is a feature. "It reads your last 30 posts and only suggests formats that already worked for you" is a mechanism. A mechanism is falsifiable: the viewer can imagine it being true or false. That is what makes the claim land instead of bounce.

Block 4: CTA

Ask for one action, phrased as an instruction, not an invitation. "Tap the link and import one project" beats "check us out." Then attach a reason the action is low-risk: a free first result, a trial, a guarantee, a deadline. A reason to act now removes "later" as the default, since "later" is where most clicks go to die. If the phrasing is not coming, a roundup of CTA examples by offer type gives you a closer that fits almost any pitch.

A fill-in ad script template

Here is the structure as a template you can paste into a doc and fill in. Keep each line inside its word budget, and read every line out loud before you keep it. If you run out of breath, the line is too long.

• Problem: "If you [specific recurring task], you already know [the annoying part]."
• Agitate: "So you [the workaround], it costs [time or money], and [the result still isn't good]."
• Turn: "Here's what nobody tells you: [the reframe, where the real cause is X, not effort]."
• Mechanism: "[Product] does [the one specific thing], so instead of [old way] you [new way] in [timeframe]."
• Proof (optional): "[A concrete, defensible detail: a number you can stand behind, a before-and-after]."
• CTA: "[Single action]. [Reason it's safe or timely]."

The Turn line is the bridge between Agitate and Mechanism. It reframes the problem from "you are not trying hard enough" to "you were solving the wrong thing," which is what makes the mechanism feel like a discovery rather than a sales pitch. It is optional, but it is the line that most often separates a script that converts from one that merely informs.

A teardown: the Dollar Shave Club script, by block

The most instructive way to learn this framework is to map it onto an ad you already know. Dollar Shave Club's 2012 launch video is a useful case because it is widely studied and the structure is visible once you look for the four blocks. The exact wording below is paraphrased to show the shape, not transcribed.

• Problem. The script opens by naming the viewer's situation directly: you are spending too much on razors with features you did not ask for. It does not open with the brand or the founder's credentials. It opens where the viewer already is, standing in a drugstore aisle annoyed at the price.
• Agitate. It makes the cost concrete and a little absurd: the twenty-dollar brand-name handle, the vibrating ten-blade contraption, the sense that you are paying for marketing rather than a shave. The cost is not just money. It is the feeling of being upsold.
• Mechanism. Here the product appears, framed as a how: a dollar a month, blades shipped to your door, no store, no markup. That is a mechanism, not a feature list. It explains why the result (cheaper, easier) is possible, which is what makes the price believable instead of suspicious.
• CTA. One action, stated plainly: stop forgetting to buy razors and subscribe. The action is low-friction and tied directly to the problem it just named.

What makes the teardown worth studying is the discipline in the Mechanism block. The ad never claims the razors are the best in the world. It claims something narrower and provable: they are good enough, and the delivery model is what is different. A claim you can defend beats a superlative you cannot. The same restraint is what keeps a narrative performance ad from drifting into a mood film that entertains and never asks for the click.

Turning one script into a tested winner

A single script is a hypothesis, and a hypothesis with no comparison is just a hope. The payoff of writing to a fixed structure is that variants become cheap to produce, and cheap variants are how you actually find the line that works.

1. Hold everything constant and vary one block. Usually the Problem opener, because the hook carries the most weight. Write three Problem lines against the same Mechanism and CTA, and you have isolated the variable that matters.
2. Ship three to five variants per round. You are not trying to predict the winner from your chair. You are giving the platform enough genuinely different angles to find it. The variant you were least sure about wins often enough that pre-judging is a mistake. A guide to what to change first in creative tests will save you a couple of wasted rounds.
3. Read the retention graph, not just the click rate. A strong click-through paired with a steep drop in the first few seconds means the hook over-promised. The retention curve is your script's report card, and you can usually map a dip to a specific block.
4. Set a kill threshold before launch. Decide the spend a variant gets before you cut it, and honor it. The script you enjoyed writing is not owed a rescue.

Common mistakes the framework prevents

• Front-loading the brand. The logo and name belong in or after the Mechanism block. The opening line is for the viewer, not the brand manager.
• A Mechanism that is secretly a feature list. If your "how it works" is bullet points of capabilities, you skipped the part that explains why those capabilities produce the result.
• A CTA with two asks. "Follow us and tap the link and use the code" splits attention. Every additional ask cuts compliance roughly in half, so pick one.
• Ignoring the muted viewer. A large share of feed video plays without sound, so a script that only works as voiceover loses half its audience the moment it autoplays. Plan the on-screen text alongside the spoken lines; the case for captioned video ads is simply that most people will read your script before they ever hear it.

FAQ

Does the four-block structure change for TikTok versus LinkedIn?

The four blocks hold across placements; the calibration shifts. TikTok and Reels reward a faster, more native Problem line and a looser tone, so you trim toward the low end of the word budgets. LinkedIn B2B video ads tolerate a slightly slower setup and more explicit stakes in the Agitate block. The efficient path is to write one strong script, then adjust the Problem line's tone and re-cut the same body for each feed rather than starting over.

Where does the proof or testimonial fit?

Proof sits between Mechanism and CTA, as a short reinforcing beat rather than its own block. It works best when it makes the mechanism more believable, not when it adds a fresh claim. A single defensible detail, such as a before-and-after or a number you can stand behind, does more than a wall of stars. If the proof requires its own explanation, it is too much for an ad script and belongs on the landing page.

How do I keep the Mechanism from sounding salesy?

Describe what the product does in terms the viewer could verify, and stop there. Replace adjectives such as "powerful," "seamless," and "all-in-one" with the specific action that produces the result. If you cannot screen-record the thing you are claiming, it is probably an adjective in disguise, and it should be cut.

The four blocks make writing scripts fast; the slow part is turning five scripts into five finished, captioned videos before the angle goes stale. If that production step is your bottleneck, Aitachyon drafts script variants in this structure and renders them so you can test a full round in an afternoon. Try it free and put your first three scripts up against each other.