Build a repeatable onboarding video series in Synthesia (that HR can update without you)

Onboarding video series fail the same way every time: they get made once, beautifully, and then rot because updating them means re-filming. Synthesia changes the failure mode — when a video is a script, updating it is a text edit. The build that survives is the one designed for the second version, not the first.

Design the series as a template before recording anything. Our set has one visual structure: avatar on the left, screen or slide on the right, chapter titles as scene breaks. Every video follows it, which means viewers learn the format once and every future video inherits the professionalism free. In Synthesia, build video one carefully, then duplicate it as the master — never start each episode from blank.

Script discipline is what makes the avatar watchable. Write for the ear: short sentences, contractions, one idea per scene. The avatar renders punctuation as delivery, so commas are breaths. And where the content shows software, put the screen recording on the right half rather than asking an avatar to describe menus — the combination of a person explaining and a screen showing is what makes it feel produced.

The update path is the reason this tool exists. When the expense policy changed last quarter, 'updating the video' was editing two sentences in the script and re-rendering — fifteen minutes, no camera, no scheduling the one person who knows the policy. Same for the German versions: the translated script renders with the same avatar, and our Berlin team gets identical training.

Two honest limits: voices have a synthetic edge on emotional content, so keep the avatar for procedural material and use a real human for the welcome video — warmth matters there. And render times mean you should batch: edit all quarterly updates in one sitting, render overnight. The series went from 'we should redo these' to 'these are current' — which is the actual goal.