- Give each chapter one step or question, and title it with the words a viewer would type to look for that moment.
- Start the first chapter at 0:00 and keep every chapter at least ten seconds long, because short chapters look like noise and get skipped.
- Check every timestamp against the finished file before you publish, since a chapter that starts two seconds early teaches the viewer to distrust the whole list.
Video chapters are the cheapest thing you can add to a video, and one of the few additions that changes viewer behavior at once. A viewer who sees a list of timestamps can judge, in a few seconds, whether the video answers their question and where to start. Without one, they either watch from the beginning and hope, or they close the tab and look for something else.
The detail is small. The effect is larger than most creators expect, and the mistakes are common enough to be worth a careful guide.
What chapters change for the viewer
A chapter list turns linear time into an index. Without one, a seven-minute video is a single block the viewer has to commit to. With one, it’s six named sections, and the viewer can see the whole shape before they commit to any of it. That reduces the cost of starting, which is where a lot of videos are lost.
Chapters also make it easy to come back. A viewer who stopped at 4:10 can return to 4:10 without remembering where they were. For tutorials that people finish over several sessions, that’s the difference between a finished video and an abandoned one. Some platforms also surface chapter moments in search results, so a chapter title can be the thing that brings a viewer in, which is one more reason to write those titles carefully.
How to choose the cut points
Cut points should fall on changes the viewer can see, not at arbitrary minutes. The natural boundaries in a screen recording are a change of page, a new decision, or a new goal. When the screen changes to a different place in the product, the viewer is about to ask a new question, and that’s where a chapter belongs.
Here’s a concrete example. A seven-minute walkthrough called “Set up client billing” might break into six chapters at 0:00 for the goal and the overview, 0:40 for creating the customer record, 2:10 for adding a payment method, 3:55 for sending the first invoice, 5:30 for handling a failed payment, and 6:20 for checking the result. Each chapter starts where the screen changes, and each one answers a question the viewer is likely to have at that moment.
Notice what isn’t there: a chapter at 3:00 that happens to fall in the middle of invoice creation. Cutting by the minute would have split a single step in two, and the viewer who wanted the invoice would have had to find the right half of it.
The diagram shows the principle. The boundaries are where the work changes, and the chapter labels tell the viewer what kind of work it is.
Branching tutorials and labeled chapters
Some tutorials branch. The setup differs between Mac and Windows, or between a free plan and a paid one, or between a team that uses single sign-on and one that doesn’t. Chapters handle branching well if you label them so the viewer can tell which path they’re on before they click.
The pattern that works is a shared opening, one labeled chapter per branch, then a shared closing. Put the branch in the chapter title and in the first words of the chapter: “Windows: install the connector” and “Mac: install the connector.” A viewer who sees their operating system in the list knows exactly where to go. A viewer on the other path skips the block, and the shared closing shows them they’ve rejoined the flow.
Avoid mixing branches inside one chapter. A chapter that switches from Windows to Mac halfway through is the kind that makes viewers rewind, and the rewind is usually where they give up. If the branches differ in only one step, a short chapter for that step with a clear title beats a long chapter that covers both.
Label the branch: put the operating system, plan or path in the chapter title itself. Viewers scan titles, not descriptions, so the label they can see is the one they use.
Video chapters: title them like a search query
A chapter title should read like the question a viewer types when they’re stuck. That means a verb and an object, with no filler. “Add a payment method” works. “Payment” doesn’t. “Part 2” is worse, because it tells the viewer nothing except that a second part exists.
Aim for three to six words. Shorter titles are often too vague to help, and longer ones are hard to scan in a list. Avoid “Introduction” unless the first twenty seconds really are an introduction. If they show the result, title the first chapter with the result, or with the goal, because the viewer wants to know what they’ll be able to do.
| Weak chapter title | Stronger chapter title |
|---|---|
| Part 2 | Create the customer record |
| Settings | Add a payment method |
| Step 3 | Send the first invoice |
| Misc | What to do when a card fails |
The stronger column reads like a list of things the viewer can do. The weaker column reads like a table of contents for an internal document, which is not how a viewer scans a video.
The timing rules that prevent broken lists
Some of the rules below come from platform requirements, and others are habits that keep a list from looking careless. The platform rules are current as of this writing, so check your host’s guidance before you rely on them.
- Start the first chapter at 0:00. Most platforms require it before they’ll show a chapter list, and it’s a sensible rule everywhere.
- Keep every chapter at least ten seconds long. At the time of writing, YouTube requires at least three timestamps, with each chapter at least ten seconds long. A chapter shorter than that looks like noise, and viewers skip it.
- Land on the frame where the new state is visible. When a viewer clicks a timestamp, they arrive at a single moment. If that moment is mid-click, with the cursor halfway to the button, they have to re-scan the screen to catch up. Start each chapter a beat before the change, so the viewer lands on the new screen and the explanation.
- Use one time format throughout. Mixing “2:10” and “02:10” looks careless even when both are correct.
Catching a cut that starts two seconds early
Chapter timestamps fail most often by a small margin. A chapter that starts two seconds before the new state appears shows the viewer the end of the previous step, and they arrive at the wrong moment. The error is easy to make, because the timestamp was typed from a rough scrub and never checked against the frame.
To catch it, check the finished file rather than the timeline you edited in. Jump to each timestamp in the published player and look at the first frame. The new state should be visible, or at least the first moment of the change. Then play back the two seconds before the timestamp. If the previous step is still on screen with the cursor moving toward the next control, the chapter is early and should move forward. If the new state is already complete a second before the timestamp, the chapter is late and should move back.
Do this check after every edit, not once at the end. A trim in the middle of a video shifts every timestamp after it, and an edit that moves one chapter often moves the ones below it. Recheck the list in the published file each time, and the list stays accurate for as long as the video does.
Build the list from the finished file
The most reliable way to write chapters is to do it after the edit, not during recording. Recording has its own rhythm, and timestamps written from memory drift. Once the video is cut, the list should come from the finished file.
- Play the finished video once, and note every point where the screen changes to a new state or the goal changes.
- Write a title for each of those points, in the verb-and-object form described above.
- Check each timestamp by jumping to it in the finished file, and confirm the first frame shows the new state.
- Remove any chapter shorter than ten seconds, or merge it into the chapter before.
- Publish the list with the video, and open the share page once to test every link.
Watch out: automatically generated chapter lists are a draft. A chapter that drifts two seconds from its scene looks careless, so verify every timestamp against the finished file before you publish.
When chapters help, and when they’re clutter
Chapters are not a universal improvement. Below roughly ninety seconds, a chapter list usually adds noise. A sixty-second video doesn’t need an index, because the viewer can see the whole thing in the time it would take to read the list. Listing three chapters in a minute-long clip makes the video feel more fragmented than it is.
The break point is where a viewer might plausibly want to skip. A short video with one outcome has nothing to skip. A longer video with several decisions, where some viewers need only one of them, is the place where chapters pay back their cost. Use the length of the video as a rough guide, but let the structure decide.
When a long video needs two levels of structure
Some videos run past twenty minutes, usually because they’re a course lesson, a full onboarding series in one file, or a recorded training session. A flat list of twenty chapters becomes hard to scan, and the viewer can’t tell which chapters belong together. The fix is two levels: a few major sections, each with its own short list of steps.
Most platforms won’t show two levels natively, so the structure lives in the chapter titles themselves. A workable convention is a prefix for the section and a colon for the step: “Setup: create the workspace,” “Setup: invite the first member,” “Reporting: build a weekly view.” The prefix groups the chapters when the list is read top to bottom, and it lets the viewer skip a whole section at a glance.
Keep the top level to four or five sections. If you need more, the video is probably two videos, and splitting it is kinder to the viewer than adding a level. The rule from the splitting discussion still applies: split on a change of outcome, and use the levels inside each part to show the steps.
Chapters and what the analytics tell you
Once a video has chapters, the analytics become more useful. If your platform reports engagement by chapter, you can see which section people replay and which they skip entirely. A chapter that’s replayed heavily is usually a step the viewer found hard, or a value they needed to copy. A chapter that’s skipped every time may be one the viewers didn’t need, or one that didn’t match what they expected from its title.
Use that information as a prompt, not a verdict. A skipped chapter could mean the title promised something the section didn’t deliver, or it could mean the section is fine and the viewers who skip it simply already know the material. The video engagement analytics guide explains how to read those patterns without over-correcting.
Chapters also need maintenance. When you re-record a section or change a step, the timestamps after it move, and a chapter title can become wrong even when its timestamp is right. A chapter called “Send the first invoice” that now opens on the invoice preview, rather than on the send button, is accurate in spirit and misleading in practice. Treat chapter titles as part of the video’s content. When a section changes, read the titles on either side of it again, and rename anything that no longer describes the first few seconds it covers.
Where to go from here
Chapters are a small job with a large return. Write each title as a question the viewer would ask, put every boundary where the screen changes, and check every timestamp against the finished file. Skip them on short videos, and use them on long ones where the viewer might want only one part.
RecordlyPro generates chapters automatically when a recording is finished, so the list is a first draft you can check in a few minutes. Open the share page, click each chapter, and confirm it lands on the screen you expect. For the length decision that comes before chapters, how long a tutorial should run explains when a video needs splitting rather than indexing.
Keep reading
Async video: stop scheduling the meeting you could record
A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.
Sofia Meyer
October 6, 20262 min readAuto-zoom in screen recordings: what it is and why your demos need it
A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.
Sofia Meyer
October 13, 20262 min readHow to choose a screen recorder for product demos in 2026
A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.
Sofia Meyer
October 16, 20262 min readDofollow vs nofollow in 2026: what still passes value
The link attributes changed, Google's treatment changed, and a lot of advice is out of date. Here's what dofollow, nofollow, sponsored and ugc actually do now.
The Recordly Team
October 10, 20262 min read