Video embed

Embedding puts a video inside a page you control — a help centre article, a product page, a knowledge base entry — usually with a small piece of HTML that loads the player in a frame. The video is hosted elsewhere and appears in place.

The alternative is uploading the file to whatever platform hosts the page, which duplicates the video, loses the analytics and makes updating it a matter of replacing files. Embedding keeps one copy and one set of numbers.

The practical questions when choosing an embed are whether the player is responsive, whether it carries branding you did not choose, and whether the viewing still reports back to the platform that hosts it. The third is the reason to embed rather than upload: a video watched in your help centre should count in the same place as one watched from a link.

Some platforms also support oEmbed, a standard that lets certain sites expand a bare URL into a player automatically. Where it is not supported, pasting a URL and expecting a player is the most common source of a broken-looking page.

Loading behaviour is worth checking on any page where the video is not the main event. An embed that loads its player immediately costs the page weight and time whether or not anybody presses play, which matters on a help centre article where most readers are there for the text. Players that load a still image first and fetch the rest on click avoid that cost.

Finally, decide where the canonical version of the video lives. A recording embedded in five places should still be one recording with one set of numbers; copying the file into each system produces five videos, five sets of analytics, and a maintenance problem the first time the content changes.