How Creators Write 30 Second Video Scripts That Actually Convert

A converting 30 second script is about 75 to 85 spoken words, structured as hook, problem, product, proof, offer, with the hook doing roughly half the work. Creators who convert consistently don’t write better sentences than everyone else. They spend a disproportionate amount of their effort on the first three seconds and treat the remaining twenty-seven as delivery.

The word count is the constraint people underestimate. Natural conversational pace runs around 150 words a minute, so thirty seconds gives you about 75 words, and most first drafts come in at 130. Everything you cut in the second pass is the reason the script works, because a script written to fit in thirty seconds sounds rushed while a script written at 75 words sounds deliberate.

Why the First Three Seconds Decide Everything

Hook rate, the share of impressions still watching at three seconds, has the widest performance spread of any metric in short-form advertising. Industry benchmarks generally put a healthy range somewhere around 25 to 35 percent, and the difference between a script that hits that and one that doesn’t is almost never the offer. It’s the opening line.

The hook has one job: name something specific enough that a viewer recognises themselves in it. “Tired of expensive skincare” fails because it’s a category, not a person. “I returned three of these before I found one that didn’t leak” works because it’s a scene, and the viewer either has that scene in their history or they don’t, which is exactly the filtering you want.

Write ten hooks before you write the rest of the script. Not two. Ten, because the first four will be variations of the same idea and the useful ones show up around seven. Then pick three and test them against identical bodies, since the hook is responsible for most of the performance variance and testing anything else first wastes budget.

The best source material for hooks isn’t your imagination. It’s your reviews, your support tickets and your comment section, where people have already written the sentence in their own words. Four-star reviews are the sweet spot: specific enough to be useful, honest enough to name a real tension.

The 30 Second Structure That Holds Up

Roughly three seconds hook, six seconds problem, twelve seconds product, six seconds proof, three seconds offer. That’s a template rather than a rule, but the proportions matter more than the exact numbers, and the mistake almost everyone makes is inflating the product section and starving the problem.

The problem beat has to use the viewer’s vocabulary, not yours. If your customers say “it kept dying halfway through the day,” your script says that, not “insufficient battery longevity.” Marketing language in the problem section is where scripts lose people, because the viewer stops recognising themselves the moment the words stop sounding like their own.

Your product beat is a demonstration with narration, not a feature list. One mechanism, one benefit, and a specific number if you have one. “Charges in 40 minutes” lands. “Fast charging technology” doesn’t, and the difference costs you nothing to fix.

Proof comes last before the offer because that’s where scepticism peaks. A review quote, a rating count, a before and after, a certification. And the offer is one instruction. One. A script that says “check the link, use the code, and follow for more” converts worse than one that says any of those three alone, which is the decision paralysis that too many options reliably produces.

Writing for the Ear Rather Than the Page

Read every draft aloud, at pace, timing it. This single habit fixes more script problems than any framework, because sentences that look fine on screen fall apart when spoken, and you’ll hear the clause you can’t get through in one breath immediately.

Short sentences. Contractions. No stacked clauses. Avoid words that require the listener to hold something in memory while you finish a thought, since a viewer in a feed has no capacity for that and will scroll rather than concentrate. The average sentence in a good short-form script runs about eight to twelve words.

Volume is what separates creators who convert from creators who occasionally get lucky, and writing twenty variants by hand is where most people quit. Plenty of creators now draft with an AI tool for writing marketing scripts to generate the first ten hook options and then rewrite the ones with promise in their own voice, which is a reasonable division of labour: the machine handles quantity, you handle whether it sounds like a person.

Whatever produces the draft, the read-aloud pass stays yours. Generated scripts tend toward even sentence lengths and slightly formal rhythm, and the fix is always the same: break one sentence in half, add a fragment, and cut the adjective doing no work.

How Scripts Change by Platform, Price and Category

TikTok tolerates a slower build than Reels does, and Reels tolerates less than YouTube Shorts, where viewers arrive from a search-adjacent mindset and give you an extra beat. Same script, but the hook can be marginally less aggressive on Shorts and needs to be sharper on Reels.

Price point changes the structure. Under about thirty dollars, the decision is emotional and the script can go hook, product, offer with barely any proof. Over two hundred, you need the proof section expanded and a call to action that sends people to read rather than buy, because a thirty second video isn’t closing a considered purchase on its own.

Category matters too. Beauty and apparel lean on visual transformation, so the script mostly narrates what’s on screen and gets out of the way. Software and B2B need the problem section to be longer and more precise, since the buyer has to recognise a workflow, not a feeling, and these often perform better at 45 seconds than at 30. Supplements, finance and anything health-adjacent are constrained by platform advertising policy before creative, so check the restricted claim rules before you write rather than after your ad gets rejected.

Audience familiarity is the variable people forget. A script aimed at cold traffic spends its budget on the problem. A script for retargeting can skip the problem entirely, open with proof, and use the whole thirty seconds on objection handling and offer, because that viewer already knows what you sell.

The habit that compounds is keeping a hook bank. Every line that outperforms goes into a file with the audience it worked on, and every review or comment with unusual phrasing goes in beside it. After six months you’re not writing hooks from scratch anymore, you’re recombining lines that have already proven they work on people who buy from you, which is a much shorter path to the next winner than a blank page ever is.