The scripting advice that dominated YouTube in 2020 told you to open with a hook, cut to a logo sting, then spend half a minute introducing yourself before the content started.
YouTube now measures exactly that stretch of your video and reports it back as a number in Studio. The six-part skeleton still works. What goes inside the first thirty seconds of it has to change.
What does a YouTube script actually have to do?
It has to win two decisions. First, whether your opening delivers what the thumbnail and title promised. Second, whether each later minute is worth staying for. YouTube’s recommendation systems watch “how long and how much of a video they watch” plus “likes, shares, comments, ‘not interested’ feedback, and survey responses”, and they use fresh performance data for each individual video rather than your channel’s track record.
Why this matters: a script is not a formality you fill in before filming. It is the only place you decide, on purpose, what a viewer gets at every minute mark.
Retention is decided in the script, not in the edit.
Because YouTube judges each video on its own data, a new upload needs its first viewers before any of that happens. Sprizzy puts the video in front of people who would choose to watch it, through YouTube’s official ad platform. The script does the rest.
What should the first five seconds of the script say?
Say the thing your title promised, in the plainest words you have. If the title asks a question, start answering it. If it promises a result, show the result. Rene Ritchie, YouTube’s Creator Liaison, framed it this way in a 2023 interview: “The first act can’t just be an introduction. The first act has to be a reward.” Write that line first, then build backwards.
Five patterns that fit in five seconds:
- The result first. Show the finished thing, the number, or the after shot, then rewind. “This cost me nine dollars. Here it is.”
- The direct answer. Answer the title’s question immediately, then spend the video earning it. Best when the title is a question someone typed.
- The contradiction. Name the thing your audience believes and say it is wrong. Only use it if the video actually proves it.
- The stake. State what happens if they get it wrong. “Set this one field incorrectly and the whole upload gets claimed.”
- The open loop. Show a moment from later in the video with no explanation. This works best when the moment is genuinely strange.
One claim to ignore: “research has shown people click away within five to seven seconds”. No one names the study. The measurable version is your retention report’s Intro segment, and it sits at thirty seconds, not five.
What is the promise line, and where does it go?
The promise line is one sentence, placed right after the hook, that tells the viewer precisely what they will have by the end. “By the end of this you will have a working template you can paste into a doc.” It is the contract. Every section after it either pays the contract off or gets cut, and the viewer can feel the difference within a minute.
Where scripts break: if your title and thumbnail promise one thing and your promise line promises something narrower, retention collapses at the moment the viewer notices.
YouTube’s own advice when your intro underperforms is to make your thumbnail and title better reflect your video content, which is the same problem seen from the other end.
The packaging and the promise line have to describe the same video.
Our guide to ranking a YouTube video in search covers the packaging half in more depth.
Should you still script a logo sting or a channel intro?
No. Cut both. A three to five second logo animation and a thirty second “hi, I’m X and on this channel we talk about Y” segment together consume the entire window YouTube measures as your Intro, and they deliver nothing the viewer came for. If you want branding, put it after the hook has landed and the promise has been made, and keep it under two seconds. Most channels are better off with none.
| Opening beat | 2020 script | 2026 script |
|---|---|---|
| Logo sting | 3 to 5 seconds, right after the hook | Cut |
| Channel introduction | Around 30 seconds on who you are | Cut, or under 2 seconds after the promise line |
| Subscribe ask | Near the 30 second mark | Late, once, after a payoff |
What changed: YouTube now reports on that exact window and tells you to experiment inside it.
A viewer arriving from a recommendation or a search result has no idea who you are and no reason yet to care, so an introduction spends attention you have not earned.
If a channel intro is load-bearing in your videos, read our list of the mistakes that keep small channels small. This one is rarely the only one.
How do you structure the main body so people stay?
Write the body as a sequence of self-contained payoffs, not as one long build. Each section should give the viewer something usable before it asks for more time. YouTube’s analytics guidance is explicit about the shape here: “If your top moments happen later in your video, consider introducing that content earlier in the video. Audience numbers tend to decrease as videos progress, so try to catch their attention earlier.”
Order by strength: your best demonstration, your biggest number or your most surprising example belongs in the first third, not saved for a finale most people never reach.
Order sections by strength, not by chronology.
If your natural structure is a story that requires chronology, put the most striking moment of that story in the hook and then tell it in order.
Chapters help the viewer skip to the part they want instead of leaving. To turn them on:
- Add three timestamps. Put at least three timestamps in ascending order in your description.
- Start at 00:00. The first timestamp is the start of the video.
- Keep every chapter at least ten seconds long. Ten seconds is the minimum YouTube accepts.
- Name them while you script. If a section is hard to name in four words, it is probably two sections.
Where do mid-video re-hooks belong?
Wherever your retention graph dips. YouTube defines dips as “moments in your video that were either skipped or moments where viewers stopped watching your video completely” and tells you to review them to understand why the audience lost interest. A re-hook is the fix: one line that restates what is still coming, placed just before the point where people leave.
Re-hooks are short and specific. “That is the setup. The part that actually saved me the money is next.” Three practical placements:
- Before any necessary boring part. Setup, disclaimers and context all cost attention. Buy it back first.
- At each dip on your last video’s graph. Same format, same audience, usually the same failure point.
- Immediately after a payoff. The moment a viewer’s question is answered is the moment they consider leaving. Give them the next question.
A viewer decides to stay again at every dip. Re-hook before each one.
How should the script end?
End on one specific next video, not a generic subscribe ask. Name the video, say what the viewer gets from it, and make it the thing they would naturally want after this one. Then script the last twenty seconds so there is still something to watch while the end screen is on screen.
Keep talking over the end screen. Silence over a static frame is where people close the tab.
Two exit tools: end screens and cards, which make the spoken next-video line clickable.
| Tool | End screens | Cards |
|---|---|---|
| Where they show | The last 5 to 20 seconds of the video | The moment you set, ideally when you reference another video out loud |
| How many | Up to four elements at 16:9 | Up to five per video |
| Limits | Video must be at least 25 seconds long; not available on videos set as made for kids | Shown to viewers on computers, so treat them as a bonus |
Once the video is live, our checklist for what to do after uploading a YouTube video picks up from here.
What does the retention graph tell you about your script?
It tells you which lines failed. The Intro figure is “what percentage of your audience still watched your video after the first 30 seconds”, and YouTube groups videos holding 50% or more into an “above typical intros” list. If you land below that, its own advice is to “edit the first 30 seconds of your video and experiment with different styles”. That is a script instruction, not an editing one.
Read the graph against your script, line by line. Key moments appear only on videos at least 60 seconds long with at least 100 views, and the data takes a day or two to process.
- Top moments. Where “almost no one left the video while watching”. Whatever you wrote there is a pattern worth repeating.
- Spikes. Rewatched or shared moments.
- Dips. Your re-hook list for next time.
- Engaged views. Since the August 2026 change to how views are counted, the gap between views and engaged views is a second read on the same problem, because engaged views count only the people who stayed past the initial seconds.
You can also test the packaging directly: Test & Compare runs up to three titles and thumbnails and gives the win to the one with the highest watch time.
What if the video has not reached 100 views?
None of this works on a video nobody watched. A hundred views is the floor for key moments, and plenty of good scripts never get diagnosed because the video never found an audience.
This is exactly what Sprizzy is for: you pick the video and who should see it, and it runs the campaign on YouTube’s own ad platform, real viewers only, no bots, from $50.
A script can only be judged by people who saw the video.
Those views are verifiable in your own YouTube Studio analytics, so the next graph you read comes from a real audience. Test it with a small budget: Sprizzy campaigns start at $50 and your first one comes with a money-back guarantee.
What does a finished script template look like?
Like this. Copy the beats into a document, fill the right-hand column with your own lines, and delete any row that does not earn its runtime. The timings assume a video of eight to twelve minutes; scale the middle, not the opening.
| Beat | Runtime | What you write |
|---|---|---|
| Hook | 0:00 to 0:05 | The title’s promise, delivered or begun. One sentence. No greeting. |
| Promise line | 0:05 to 0:15 | What the viewer will have by the end, stated plainly. |
| Context | 0:15 to 0:30, optional | Only what the viewer cannot follow the video without. Cut it if in doubt. |
| Section 1 | 0:30 onward | Your strongest payoff. Not your first chronologically, your best. |
| Re-hook | Before each dip | One line naming what is still coming. |
| Sections 2 to 4 | Middle | One self-contained payoff each, in order of strength. |
| Comment prompt | Around the midpoint | One specific question tied to the content, not “let me know below”. |
| Final payoff | Last third | The result, the summary, or the thing you teased in the hook. |
| Outro | Last 20 seconds | One named next video and why they want it. Keep talking over the end screen. |
| End screen | Last 5 to 20 seconds | Up to four elements. The next video first, subscribe second. |
Write the hook, the promise line, every re-hook and the outro word for word. Work from bullet points through the middle so the delivery stays natural. The scripted lines are the ones the retention graph judges hardest.
How is a Shorts script different?
Compress everything and drop the ending furniture. A Short can run up to three minutes, but there is no room for context and the payoff has to arrive in the first few seconds. The structure is three beats:
- One promise. The hook and the promise line in a single opening sentence.
- One payoff. Delivered in the first few seconds, not the first third.
- One reason to watch it again. Rewatches are the Short’s version of a spike on the graph.
The ending changes too. End screens need a video of at least 25 seconds, and cards show for viewers on computers, so neither is your route out.
Instead, add a related video link, which appears as a clickable link below your channel handle and requires advanced feature access. Write the last line of the Short so it points at that video specifically.
A Short is one promise, one payoff and one reason to watch again.
Because key moments only appear on videos of at least 60 seconds, shorter Shorts are diagnosed by their engaged view ratio rather than by a retention graph.
Common questions
How long should a YouTube intro be?
As short as you can make it, and ideally there is nothing you would call an intro. YouTube measures the Intro segment of your retention graph at the 30 second mark, and videos holding 50% or more of their audience past that point are grouped as above typical. Spend those 30 seconds delivering, not introducing.
Should I put a channel logo animation at the start of my video?
Not at the start. A logo sting spends the most valuable seconds of the video on something the viewer did not come for, and it lands inside the window YouTube reports back as your Intro figure. If you want the branding, place it after the hook and the promise line, keep it under two seconds, and check the retention graph to see what it cost.
Do I need to write a word-for-word script?
Only for the parts where the exact words matter. Most creators write the hook, the promise line, every re-hook and the outro word for word, then work from bullet points through the middle so the delivery stays natural. Those scripted lines are the ones the retention graph judges hardest, so they are worth the extra time.
Where should the subscribe ask go in a script?
Late, and only once. The old advice put a subscribe ask around the 30 second mark, which spends your entire Intro window on a request rather than a payoff. Ask after you have delivered something, tie it to what the viewer just got, and put one specific next video ahead of the generic ask in your outro.
How many cards and end screen elements can I add?
You can add up to five cards to one video, and up to four end screen elements on a standard 16:9 video. End screens need a video at least 25 seconds long and sit in the last 5 to 20 seconds. Neither is available on videos set as made for kids, and cards are shown to viewers on computers.
Do chapters improve retention?
They give a viewer who would otherwise leave a way to skip to the part they want. To turn them on, put at least three timestamps in ascending order in your description, start the first at 00:00, and make every chapter at least ten seconds long. Write the chapter names while scripting, since a section you cannot name in four words is usually two sections.
How do I script a YouTube Short?
Merge the hook and the promise into one opening line, cut the context entirely, and deliver the payoff in the first few seconds. A Short can run up to three minutes, but the structure is one promise, one payoff and one reason to rewatch. End screens and cards are not your exit route, so add a related video link instead.
How do I tell whether my script or my thumbnail is the problem?
Look at where people leave. If viewers arrive and drop inside the first 30 seconds, your packaging and your opening disagree, and YouTube's own advice is to make the thumbnail and title better reflect the video. If they stay past the intro and leave at a specific later point, that is a script problem and the dip names the line that caused it.
