Chapters that let a viewer jump between sections of a tutorial can live in three different places. Which one is used determines whether they survive an upload.
Containers can carry chapters natively
Common video containers support a chapter track, which is a list of timestamps and labels stored alongside the video and audio streams inside the same file.
Local players read that track and present a chapter list. Nothing about the picture changes, and the file size increase is negligible because chapters are only text and timing.
The limitation is that this track is a property of the file, and most upload pipelines discard everything they do not need when they re-encode.
Platform chapters are parsed from text
Most video platforms derive chapters from timestamps written into the description field rather than from anything inside the file itself.
The parser looks for lines beginning with a time in minutes and seconds, requires the first one to start at zero, and expects the list to run in ascending order.
A minimum count and a minimum length per chapter are usually enforced, which is why a tutorial with many very short steps often shows no chapters at all.
Sidecar files sit between the two
Editing and transcoding tools frequently export chapters as a separate text file, listing each start time and title on its own line.
This format is convenient because it can be edited without touching the video, versioned alongside a project, and converted into either an embedded track or a description block.
For anyone producing a long tutorial series, keeping chapters as a sidecar is the most durable option, since the same list can be reused across every destination.
Chapters change how a tutorial is measured
Once chapters exist, viewers stop watching linearly. They jump to the section they need, which fragments the retention curve into a series of separate entry points.
That is usually the correct outcome for reference material, but it changes what the numbers mean. Average view duration falls even when the video is serving its audience better.
Creators who compare a chaptered tutorial against an unchaptered one on retention alone will draw the wrong conclusion about which format works.
Where chapters and accessibility overlap
Chapter titles are text associated with a time range, which makes them useful to the same systems that consume captions and transcripts.
Descriptive chapter names help a search system understand what a section covers, so a name like adjusting the exposure triangle carries more weight than part three.
Writing chapters as short descriptive statements rather than as numbered labels is therefore worth the effort, independent of what any particular platform does with them.