How to Write Short-Form Video Scripts Fast
The short-form script procedure senior creators run instead of staring at a blank doc. Anchored to Jenny Hoyos's 60-second discipline, the MrBeast leaked production handbook, Adam Mosseri's January 2025 ranking framework, and Kendall Hope Tucker's Ramp brand-twist rule. Includes the hook-bank, the 3-block timing template, and the 10-minute draft sprint.
By Bell Chen, founder. Last updated May 19, 2026.
Jenny Hoyos, the YouTube Shorts creator profiled by Marketing Examined (marketingexamined.com) who has been one of the most-studied 60-second video operators of the last three years, named the script-writing discipline that keeps her cadence at three to five Shorts per week: separate writing days, filming days, and editing days into non-overlapping blocks, and treat the script as the load-bearing deliverable of the writing block. The follow-on rule in Hoyos's working method, drawn from the same Marketing Examined writeup: the hook is not the first sentence of the script, it is the entire structural decision the script gets built around, and a script with a weak hook cannot be saved by strong production.
Most write a short-form script in 10 minutes guides invert that order. They give a template that pre-allocates 5 seconds of intro, 30 seconds of body, and 10 seconds of CTA, and they treat the hook as a one-line field to fill in. That order produces scripts that ship on time and never break out. This page is the procedure for replacing the fill-in-the-blanks ritual with a hook-first sprint that researches the hook bank, picks the strongest hook against the platform-rewarded signal, and writes the rest of the script downstream of the hook decision. The end-to-end procedure runs in 45 to 60 minutes for the first script and 10 to 15 minutes for subsequent scripts in the same arc, once the hook bank is built. It is the procedure I use on the small B2B product account I operate.
What You'll Need
- A clear brand or content pillar the scripts will be drafted against
- A calendar block of 2 hours for the one-time hook-bank build
- Access to your last 30 to 60 days of analytics so the hook bank can be sorted by signal target
Time: 45 to 60 minutes for the first script, then 10 to 15 minutes per script after the hook bank is built
What we're actually solving
The thing most fast-script pages get wrong is treating the hook as the opening line. The hook is the decision about what tension the entire 30 to 60 seconds resolves. Once the hook decision is locked, the body writes itself in 8 to 12 minutes because every line has a known target. Without the hook decision locked first, the body sprawls and the script consumes 45 minutes producing a draft that gets discarded because the hook does not survive the read-aloud test.
Three things have hardened in the short-form-script conversation since 2023. The platform-side ranking signals are public: Adam Mosseri's January 8, 2025 Reel on @mosseri (instagram.com) named the three Reels ranking signals as "Watch time, likes per reach, and sends per reach," per Mosseri. The script's hook is the single most important determinant of which signal the script ends up exercising. A tutorial hook exercises watch time. A named-number contrarian hook exercises sends. An aesthetic-storytelling hook exercises likes per reach.
The competition density is high: median engagement rates have dropped sharply year-over-year across platforms. Scripts whose hook lands in the top 20% of the hook bank typically reach 3x to 6x the median; scripts whose hook lands in the median produce median reach. The math says hook quality compounds more than script quantity.
The cadence floor is 3 to 5 publish-ready scripts per week: that band is the one most reliably modelled by the algorithms across platforms for accounts under 100,000 followers. What writing a fast short-form script is not. It is not starting with a blank Google Doc and a coffee. It is not borrowing a template and filling in product features. It is not asking ChatGPT to write me a TikTok script about my product.
Step by step
- 01
Step 1. Build the hook bank (90 minutes one-time, 15 minutes per week to refresh)
Open a single document and list 30 to 50 candidate hooks, organised by hook type. The hook types that have held across the 2026 short-form audits I have run are: named-number ("This $32 serum has 80% of what's in the $200 one"), named-person ("Why Cassey Ho's new content does the opposite of what you'd expect"), contrarian-statement ("Everyone says you need to post daily. They are wrong. Here's the math."), question ("Why do most B2B SaaS demos lose the buyer in the first 90 seconds?"), demo-as-hook (the first 2 seconds is the product doing something visually surprising, no spoken hook required), trend-audio (a trending sound paired with an on-screen text hook that names the brand-specific twist), and reaction (responding to a category news moment or competitor post). Each hook in the bank should fit on one line and name the specific tension. Spend 90 minutes the first time; refresh weekly with 5 to 10 new candidates.
Deliverable
A 30-to-50-entry hook bank document organised by hook type and signal target.
- 02
Step 2. Pick the hook against the signal target (5 minutes per script)
For each script, decide the signal target first (watch time, likes per reach, or sends per reach, per Mosseri's three-signal framework). Most weeks the calendar already names the signal target for each slot, so this step is a lookup, not a fresh decision. Then pick the hook from the bank's matching section. Tutorial pillar with watch-time target: pick a demo-as-hook or a named-number explainer hook. Commentary pillar with sends target: pick a contrarian-statement or named-person hook. Aesthetic pillar with likes-per-reach target: pick a named-number or trend-audio hook with the brand-twist layered.
Deliverable
A locked hook line plus the signal target written at the top of the script.
- 03
Step 3. Write the 3-block draft in 10 minutes (10 minutes per script)
The 3-block timing template is Hook (0 to 3 seconds), Body (3 to 50 seconds), Payoff plus CTA (50 to 60 seconds). The Hook block is the hook line from Step 2, paired with a visual cue that the camera will execute. The Body block has three to four beats, each one a single sentence that earns its place by either advancing the named tension or paying off a sub-promise. The Payoff plus CTA block names the resolution to the hook's tension and the next-video CTA. The draft fits on half a page; if it overflows to a full page, the Body has crept beyond four beats and the script is now a 90-second draft pretending to be a 60-second script.
Deliverable
A half-page draft with Hook, Body (3-4 beats), and Payoff-plus-CTA blocks.
- 04
Step 4. Apply the read-aloud pass (5 minutes per script)
Read the script aloud at the pace you will actually deliver it on camera. Two failure modes show up. The script is too long: the read-aloud pass finishes at 75 to 90 seconds because the Body has too many beats. The fix is cutting one full beat, not trying to talk faster. The hook does not land: the read-aloud pass produces a flat first sentence because the hook was read silently in your head with implied emphasis that does not survive the spoken delivery. The fix is rewriting the hook line to carry the emphasis explicitly in word choice and cadence. Roughly 30 to 40 percent of first-draft scripts fail the read-aloud pass.
Deliverable
A revised script timed at 55 to 60 seconds on natural-pace read-aloud.
- 05
Step 5. Run the MrBeast three-question gate (3 minutes per script)
The gate from the leaked MrBeast internal production handbook (simonwillison.net), the 36-page document published in September 2024 and widely cited as the most explicit production system from the highest-output short-form-graduating-to-long-form creator of the era, names the hook as "the most important part of the video," per the handbook. Question 1: does the hook name a specific tension in the first 3 seconds? If no, rewrite the hook. Question 2: is the payoff promised in the first 5 seconds? If no, add a payoff-promise line into the hook block. Question 3: can the viewer reasonably predict what the next 50 seconds will deliver? If no, the body is meandering and the structure needs a sharper through-line.
Deliverable
A three-checkbox gate marked pass/pass/pass before the script ships to the production block.
- 06
Step 6. Tag the script with its two named numbers (2 minutes per script)
The measurement rule is to pick the two or three numbers that would actually change what you do tomorrow. Before the script leaves the writing block, name the two metrics it is expected to move. For a sends-target script, the two numbers are typically 3-second retention and sends per reach. For a watch-time-target script, they are average watch percentage and saves per reach. The named numbers travel with the script into the filming and editing blocks so the camera operator and editor know which beats are load-bearing for which signal.
Deliverable
Two named numbers written at the top of the script alongside the hook line.
What good looks like
Five named benchmarks anchor what a healthy script-writing workflow should produce. First, the hook bank stabilises at 25 to 35 strong, brand-twisted entries by week 12, refreshed monthly. Second, the script-writing time per script falls to 8 to 12 minutes after the bank lookup, not the 45 to 90 minutes it takes from a blank doc. Third, 3-second retention lifts by 8 to 15 percentage points across the new scripts by week 4. Fourth, sends per reach on the commentary cluster lifts by 0.2 to 0.6 percentage points by week 4.
Fifth, reach typically sits 30 to 70 percent above the pre-procedure baseline by week 12, concentrated in the sends-driven commentary cluster where the brand-twist discipline produces the largest gap from generic-hook competitors. The Jenny Hoyos working rule that holds the system: separate writing days, filming days, and editing days into non-overlapping blocks. A workflow that tries to write, film, and edit on the same day collapses around week 3 regardless of how good the hook bank is.
Common mistakes
I have skipped the hook-bank build because I told myself I would do it next week. After 8 weeks the script quality had not improved and the writing block had expanded from 2 hours to 4 hours because every script was a fresh hook search. The fix is the Step 1 90-minute one-time build, blocked as non-negotiable in the first week. Without the bank, the rest of the procedure does not function.
I have chosen the hook after writing the body. I drafted the script with a draft-only hook line, intending to come back to the hook at the end. The body was structured around a different unstated tension than the hook eventually used, so the script read disjointed and the hook did not actually set up the payoff the body delivered. The fix is the Step 2 hook-first rule: lock the hook before writing line one of the body.
I have let the script overflow to 75 to 90 seconds. Most common failure on first drafts. The Body block stretched to five or six beats, the read-aloud landed at 82 seconds, and I tried to fix it by talking faster on camera. The fix is cutting one full beat, not the delivery pace. A 90-second script forced into 60-second delivery loses its hook and its payoff.
I have shipped scripts whose brand-twist was missing or generic. The hook read as one any account in the category could run. The Tucker brand-twist rule was skipped. The fix is the Step 1 bank refresh: every hook gets a brand-specific clause that competitors cannot copy without inheriting the brand's specific authority.
Metrics to track
3-second retention rate per script. Pulled from Instagram Insights or TikTok Creator Center. Target: 80% or above on watch-time-target scripts, 85% on sends-target scripts.
Sends per reach. The leading indicator for the sends-target scripts. Floor: prior-90-day cluster baseline; target: 0.5 to 0.9 percent on Reels per Socialinsider's 2026 industry data.
Script-writing time per script. Target: 8 to 12 minutes after the hook bank stabilises (week 12+).
Hook-bank refresh cadence. Target: 5 to 10 new candidates added per week; 30 to 40 percent of original bank entries replaced over the first 12 weeks.
Read-aloud failure rate. Target: 30 to 40 percent of first drafts requiring a rewrite. Below 20 percent suggests the read-aloud pass is being skipped or run too lightly.
Where a planning-first tool fits
The procedure runs in a single document plus a stopwatch plus the script-tracking sheet from the monthly audit. None of those steps require a third-party tool. The places a tool earns its slot are bookkeeping-heavy: pulling the candidate hooks from your competitor analysis and audience comments without manual copy-paste, scoring the hooks against the named-number and contrarian patterns the data says actually move sends per reach, and routing the script's named numbers into the same monthly audit the calendar uses. A planning-first tool that takes the brand profile as input and outputs candidate hooks pre-sorted by signal target is one option among several, alongside a spreadsheet plus a notes app plus a stopwatch. Superdirector is the planning-first tool I built around this procedure.
Disclosure by Bell Chen, founder of Superdirector: the planning-first tool referenced above is the product I build. The procedure on this page is what I use on the small B2B product account I operate, and a structurally identical version is what I have watched friends-of-the-house creator-led operators run in 2026. The Vespera Skin worked example in the source draft is a fictional composite calibrated against 2026 cross-brand benchmarks.
Frequently asked questions
Why is the hook bank the load-bearing step instead of the writing template?
The hook is the decision; the rest of the script is the implementation of the decision. Templates without a hook bank produce scripts that ship on time and never break out, because every script reaches for a fresh hook under time pressure and lands the median. A pre-built bank, refreshed monthly with the patterns that out-performed, raises the floor on hook quality and turns the 10-minute draft sprint into an exercise in execution rather than invention.
How long should a short-form script actually be?
For a 60-second Reel or Short, the script fits on half a page (roughly 110 to 140 words spoken at natural pace). For a 30-second Reel, the script is roughly 60 to 80 words. For a 15-second Reel, 30 to 40 words.
Should I write a separate script for TikTok, Reels, and Shorts, or one script that posts everywhere?
The script can be shared, but the hook should be tuned per platform if the signal targets differ. Reels and Shorts favour sends-driven and watch-time-driven content with similar hook structures; TikTok favours faster-cut, more reactive hooks. The simplest rule for an account running all three: write one script with three hook variants, each tuned to the platform, with the body content shared.
What about scripting AI-generated UGC or avatar content?
The same hook-first procedure applies. AI UGC and avatar scripts are usually written tighter than founder-on-camera scripts because the avatar cannot improvise; every line has to be locked. The hook bank is more important for AI UGC than for organic content because the avatar lacks natural emphasis.
Should I use a script template I can paste into Notion or ClickUp?
A template is fine for the 3-block timing layout. The template does not replace the hook bank, the hook decision, or the read-aloud pass. Use the template for the body; build the bank separately and treat it as the unskippable input.
How do I write a script when I do not have a clear product or brand to position?
The brand-twist rule still applies, but the twist becomes the operator's own perspective or domain authority rather than a product feature. A chemistry PhD who tests skincare claims has a credential twist. A consultant has a portfolio twist. The Tucker rule is about authority competitors cannot replicate; the source of the authority is flexible.
What is the single highest-leverage upgrade to most short-form script-writing workflows?
Building the hook bank as a one-time 90-minute investment, sorted by hook type and signal target, refreshed monthly with the patterns from the scripts that out-performed. The change converts the script-writing block from a hook-search exercise into a hook-execution exercise. The reach lift across the 12-week mark is large because the hook is the single most important determinant of which Mosseri signal the script ends up exercising.
Start with your brand, product, profile, or video
Build your hook bank with brand-profile and signal-target analysis
Generate a campaign brief