To write a YouTube script, start from the promise your title and thumbnail make and build every section around keeping it: confirm that promise in the opening, tell viewers how the video is organized, deliver the payoff section by section and close with one clear next step. Then test the draft against what viewers actually did. YouTube's audience retention report shows the share of viewers still watching after the first 30 seconds, along with the dips, spikes and top moments that reveal which lines held attention and which lost it.
Below you will find a fill-in template with a retention-check column, a worked opening and a review routine for after you publish. This is about the words. When you need to plan every frame, pair the script with the video storyboard template.
What YouTube says your opening has to do
YouTube's guide to content performance in its recommendation system calls the first seconds the point where viewers decide to stay or move on, for long-form videos and Shorts alike. It asks for an intro that delivers on the title and thumbnail, stays concise and gets to the value quickly, followed by storytelling that builds anticipation and curiosity. The scripting rule that follows is simple: settle the packaging before you write line one. If you cannot say what the title and thumbnail promise, you cannot write a hook that confirms it. Our YouTube SEO guide covers writing that package for search.
The retention report puts a number on the opening. On the key moments card, the intro figure is the percentage of viewers still watching after the first 30 seconds, and YouTube's suggested fixes are to make the title and thumbnail reflect the video better and to experiment with those opening seconds. The Content tab tips add that videos with 50% or more of the audience still watching at the 30-second mark appear in an above-typical intros list.
The five parts of a retention-first script
- Hook. One or two lines that show viewers they are in the right place: name the problem, show the end result, or open with the most surprising true thing in the video.
- Promise. What viewers will have by the end and, briefly, why you are a credible guide. Keep credentials to a sentence; the payoff is the proof.
- Structure. A one-sentence roadmap of the sections, written to match your chapter titles so the description, the chapters and the spoken words agree.
- Payoff. The body, one question per section. Lead with your strongest material: the retention help page notes that audiences shrink over a video's length, so top moments buried late are worth moving earlier.
- Next step. One call to action that fits what the viewer just watched, such as a related video or a question for the comments, rather than a list of requests.
Fill-in YouTube script template
Copy the table into your document. Fill the first three columns before you film, write your planned timestamps as you rehearse, and leave the last column empty until the retention data arrives.
| Section | What to write | Timestamp | Retention report check | What to change next time |
|---|---|---|---|---|
| Hook | The problem or result, stated in the words of the title | Planned start | Intro figure on the key moments card, and any dip right after the hook | The hook line you would rewrite |
| Promise | What viewers will have by the end and why to trust you on it | Planned start | Whether the curve levels out once the intro ends | Credentials to trim or move |
| Roadmap | The order of sections in one sentence, matching the chapter titles | Planned start | A dip here usually means the roadmap runs long | Words to cut from the roadmap |
| Section one | The most valuable or most requested part of the answer | Chapter start | Top moments here mean the order is working | Material to bring forward |
| Section two | The next question viewers will have, answered directly | Chapter start | A gradual decline is normal; a sharp dip points at a specific line | The line at the dip, quoted |
| Re-hook | A preview of what is still coming, placed before a natural exit point | Planned start | Does the curve flatten after it? | Keep, move or drop the re-hook |
| Section three | The step, example or demonstration that completes the promise | Chapter start | Spikes: rewatched for value, or because it was unclear? | Steps to slow down or show on screen |
| Recap | The answer restated in a line or two | Planned start | How many viewers leave before the recap begins | Whether the recap earns its place |
| Next step | One call to action and the video your end screen will show | Final seconds | End screen element click rate in the Engagement tab | A better end screen pick |
A worked example: the opening of a how-to video
Here is the template filled in for an illustrative video titled “Why your voice sounds hollow on camera and how to fix the room.” Notice that every line either confirms the title or tells viewers what happens next.
- Hook: “If your voice sounds hollow when you play it back, the fix may be the room you record in, not the microphone you bought.”
- Promise: “By the end, you will know how to make a small room sound drier with things you already own, and how to check the result before you film.”
- Roadmap: “First, where to sit. Then what to hang or move. Last, a quick test recording so you can hear the difference.” The chapter titles become Where to sit, What to hang or move and The test recording.
- Re-hook before section three: “Everything so far is free. The test recording is where you find out whether it worked.”
- Next step: “If your recording still sounds off, the next video on screen covers microphone placement.”
Each line is short enough to say in one breath, and none of them asks the viewer to wait for the point.
Write for the ear, then mark up for the edit
- Read every line aloud. If you run out of breath, split the sentence; if you stumble, the viewer will too.
- Keep one idea per sentence and signpost each turn, for example “That is the setup. Here is the part most people skip.”
- Add a second column for visuals: B-roll, on-screen text and graphics. Shot-by-shot detail belongs in the storyboard, not the script.
- Write chapter titles into the script. The chapters page needs timestamps starting at 00:00, at least three of them, each chapter at least 10 seconds long.
- Bring the script to the shoot. YouTube's filming tips suggest keeping copies of your talking points, such as an outline or script, and a call sheet for bigger productions.
- Reuse the final script for captions. The captions page lets you paste a transcript for auto-sync, provided it is in the spoken language, and does not recommend this route for videos over an hour or with poor audio. Edit the transcript wherever you ad-libbed.
How long should the script be?
As long as the promise needs and no longer. YouTube's notes on its recommendation system say there is no universal ideal length, advise against filler and artificial stretching, and point to retention curves to see whether viewers leave early, at specific points or stay to the end.
Estimate runtime with a timed read-through at your natural speaking pace rather than a word-count formula. Then go section by section and ask whether each one moves the viewer closer to the promise. A section that only proves how much you know is the first cut.
Calls to action that do not cost you the ending
Script the close around the end screen. The end screens page says they can fill the last 5–20 seconds, hold up to four elements on a standard 16:9 video and require a video at least 25 seconds long; viewers can hide them, and they are unavailable on videos made for kids. Write your last lines so they point at the video you will place there, and leave the screen space clear.
For subscriptions, YouTube's fanbase tips suggest choosing moments where a reminder fits and explaining that the bell lets fans pick all, personalized or no notifications. Asking is fine: the spam policy says it is okay to ask viewers to like, comment or subscribe, but offering rewards in exchange for likes, views or subscribers counts as engagement manipulation.
In a Short, the handoff is the related video link, which appears below your channel handle and needs advanced feature access. Script the last line of the Short to make that link the natural next step.
Check the script against the retention report
Run this review on every scripted video and write the findings into the last column of the template.
- Give the data time: the retention help page says it typically takes 1–2 days to process.
- In YouTube Studio, open Content, choose the video, select Analytics and find the audience retention report on the Overview or Engagement tab.
- Compare the curve with typical retention, which the same page describes as a comparison with your 10 latest videos of similar length.
- Read the key moments. That page says they appear only on videos at least 60 seconds long with at least 100 views, so a newer channel may need to read the curve shape instead: flat, gradual decline, spike or dip.
- Find the script line at the timestamp of each dip and quote it in the template.
- Judge each spike: a section worth expanding into its own video, or a passage viewers replayed because it was unclear.
- Switch segments to compare new and returning viewers, and subscribers with non-subscribers, to see whether the opening works for strangers as well as fans.
- Carry one or two changes into the next script instead of rewriting your whole style at once.
YouTube Studio can also review drafts. The Ask Studio help page says you can paste a script for feedback based on creative best practices, or select a draft, private or unlisted video for suggestions informed by your historical retention data. The same page warns that response quality varies, that answers do not reflect YouTube's views, and that conversations are deleted after 30 days, so do not paste anything confidential. The older Inspiration tab, which offered hook and outline suggestions, is being retired: its help page says the gradual deprecation started in August 2026 and points creators to Ask Studio.
Scripting Shorts is a different job
A Short has no room for a roadmap. The first frame is the hook, the body is a single idea and the ending hands off to something longer. The intro metric was built for long-form viewing; for Shorts, the Reach tab offers a stayed-to-watch metric, the percentage of times viewers kept watching past the opening seconds of a Short.
Judge Shorts scripts against other Shorts. YouTube's analytics tips warn against comparing performance across formats, because the people who enjoy your long videos may not be the ones watching your Shorts. For how Shorts can feed long-form viewing, see our guide to growing a YouTube channel, and for how the Shorts feed differs from Home and search, the YouTube algorithm explainer.
Limitations of a retention-first script
The retention report shows when viewers left, not why. A dip can come from audio, pacing or an edit as easily as from the wording, so treat the script as one suspect among several. Smaller videos may never reach the thresholds for key moments, and a single video's curve is a thin basis for changing your whole approach.
Holding attention is also not the whole goal. YouTube's recommendation guidance treats satisfaction as a separate signal from engagement, and a script tuned only to keep people watching can drift into teasing instead of delivering. No template supplies subject knowledge, accuracy or the permissions you need to film, and AI feedback on a draft is a suggestion to weigh, not a verdict.