Chapters, Cards, and End Screens: Small Levers That Add Up

Chapters, cards, and end screens are the parts of a video most creators set up once, copy from an old upload, or skip entirely. None of them will turn a flop into a hit. But each one is a small lever that nudges a viewer toward the next thing instead of toward the exit, and the rules around them are specific enough that getting them wrong wastes the lever completely.

This is the practical version: what each feature is, the exact constraints YouTube enforces, and when it is worth the few minutes of setup.

#Chapters: navigation, with strict rules

Chapters split your video into labeled sections on the progress bar. You create them by adding timestamps to your description, and the requirements are non-negotiable. YouTube needs the first listed timestamp to be 00:00, at least three timestamps in ascending order, and each chapter to run a minimum of 10 seconds. Miss any one of those and chapters silently fail to appear.

There is also an automatic version. YouTube can generate chapters for you, but your manual timestamps always override the auto-chapters. If you care about how the sections are labeled, write them yourself. The labels are short pieces of text the viewer reads, so treat them like mini headlines rather than "Part 1, Part 2."

Chapters mostly help longer, reference-style videos: tutorials, reviews, anything a viewer might want to jump around in. On a tight three-minute video there is nothing to navigate, and chapters just add clutter. Match the feature to the format.

#Cards: in-video nudges, capped at five

Cards are the small clickable prompts that appear in the top-right corner during playback. You can add up to five cards per video, and they come in a few types: Video, Playlist, Channel, and Link. The Link card, which points to an approved external website, is restricted to channels in the YouTube Partner Program. Cards are also not available on content marked as made for kids.

The discipline with cards is timing. A card that pops in the first ten seconds invites a viewer to leave before you have given them a reason to stay. The better play is to place a card at the moment you reference something, so a card to a related video appears exactly when you say "I covered this in another video." That turns a card into a relevant offer instead of an interruption.

#End screens: the last twenty seconds

End screens are the clickable elements that fill the final stretch of a video: another video, a playlist, a subscribe button, or an approved link. They can appear in the last 5 to 20 seconds of a video, the video has to be at least 25 seconds long to have one, and you can place up to four elements on a standard 16:9 video.

End screens matter more than cards because they sit at the exact decision point: the viewer just finished, and they are about to either watch another of yours or drift off. A strong end screen pointing to a genuinely related next video is one of the cleanest ways to keep someone inside your channel, which is the whole game for session time. The catch is that you have to design your ending to leave room. If your video hard-cuts to black the instant you finish talking, there is nowhere to put the end screen.

Feature	Limit	Key rule
Chapters	At least 3 timestamps	First must be 00:00, each chapter 10s or more
Cards	Up to 5 per video	Link cards are YPP only; none on made-for-kids
End screens	Up to 4 elements	Last 5 to 20 seconds; video must be 25s or longer

#How these feed discovery and retention

These features are not ranking hacks. Chapters help with navigation and can make your content easier to understand, cards and end screens help with what YouTube measures as session time, the idea that the platform rewards keeping viewers watching across videos, not just within one. End screen clicks even show up as their own line in the Reach tab, separate from suggested videos and browse. If you want the full breakdown of where each kind of view comes from, we mapped it in your traffic sources, decoded.

So the honest framing is modest. A good end screen will not make a bad video succeed, but across a channel that keeps viewers moving from one video to the next, it compounds. That is why the home and suggested surfaces, which lean on watch history and the current video, tend to favor channels that hold attention. We get into that machinery in how the home feed works.

#Reading these moves on other channels

Once you set these up well on your own videos, you start noticing them on everyone else's. Which video does a competitor send their end screen traffic to? It is almost always the one they most want viewers to see next, which tells you what they think their best or most strategic upload is. A channel that suddenly starts cross-linking an old series in its cards is reviving that series on purpose.

The bigger signals, though, are the packaging changes around the video itself: a fresh thumbnail on an old upload, a rewritten title, a test running in the background. Those are the moves that reveal what a channel is actively working on, and they are easy to miss if you only see each video once.

#The five-minute setup that pays off

You do not need to obsess over these. A sensible default per upload looks like this:

• Add chapters to longer reference videos only, with 00:00 first and clear labels.
• Drop one or two cards at the moments you actually reference other content.
• Always set an end screen, and design your ending with room for it to live.
• Point the end screen at a genuinely related next video, not just your newest one.
• Skip the features that do not fit the format instead of forcing all of them.

These are small levers, and they behave like small levers. Used consistently across a channel that already makes things people want to watch, they keep more viewers inside your world for longer. Used on a video nobody clicks, they do nothing, because the click has to come first.