Knowledge Bank

3 min read

The chapter generator, and the rules YouTube enforces silently

Three rules decide whether your chapters render at all, and breaking one fails the whole list without warning.

Not sure if Package is your weak phase? The free 3-minute workflow audit will tell you.

Flat illustration of a segmented progress bar.
PackageLinked task: Add the chapters · step 29 of the 47

Add the chapters is step 29 of the 47, and it is a four-minute job that most people skip because the four minutes involve sitting through your own video with a notepad. The chapter generator does the tedious half. It reads the caption track and hands back a timestamp list you can paste into the description.

Paste the URL
It reads the captions
Boundaries snap to sentences
You rewrite the labels

The rules YouTube enforces without telling you

Chapters render only if the list obeys three conditions. The first chapter starts at 0:00, there are at least three of them, and none is shorter than ten seconds.

Break any one of those and the whole list silently fails. No error, no warning. The description just renders as plain text and the progress bar stays unsegmented.

That silent failure is why so many creators believe chapters are broken or unavailable to them. Usually the list was one rule off. The generator applies all three before it gives you anything, which removes the most common way this goes wrong.

What it does with the timings

It spaces chapters across the runtime, then moves each boundary onto a sentence break so a chapter does not open mid-thought. A chapter that begins three words into a sentence reads as a mistake even when the timing is technically fine.

This is the part worth having done for you. Finding sentence boundaries by ear means scrubbing, and scrubbing your own video is how a four-minute job becomes an hour.

The labels are not finished

What you get backState
The timestampsDone. Accurate to the caption track and rule-compliant
The chapter labelsA first draft, lifted from what is said at that moment

The labels come straight from the words spoken at each timestamp, which makes them accurate and completely flat. "So the next thing is the settings" is a true description of that section and nobody has ever clicked it.

Rewrite them. A chapter label is a mini title competing for a skim, and the people reading it have already clicked, so they are the most valuable audience you have. Name the outcome in each section rather than the topic.

When to use it, and when not to

Chapters earn their place on tutorials, reviews, long talky videos and anything somebody might return to for one specific part. They are worth less on a story-led video with a payoff, because a labelled list is a spoiler with timestamps attached.

They also will not rescue a video nobody clicks. Chapters improve the experience of a video that already got the click, and they can surface individual moments in search, which is a real second door in. They do nothing for the first door.

Why yours might not work

The video needs captions, manual or auto-generated, because that is the only place the timings can come from. A video with captions disabled has nothing to read.

Very short videos fail too, and correctly so, since three chapters of at least ten seconds cannot fit into a runtime that short.

Keep reading

The full case for chapters, including when to leave them off, is in the chapters guide. The rest of that box is covered in writing the description, and if you want to rewrite the labels properly, the thinking in picking the title applies to every one of them.