A document and a voiceover do different work. Documents can hold detail. Voiceovers create momentum. If the narration repeats every word already on the page, the viewer is forced to read and listen to the same information at the same time.
The script should make the document easier to understand. It should explain the thread between pages, focus attention, and add context the screen cannot carry gracefully.
Extract the argument first
Read the PDF once without writing. Then summarize it in five to seven sentences. Those sentences are the backbone of the video, not the page titles.
Ask:
- What does the audience know at the beginning?
- What do they need to understand next?
- What evidence changes their view?
- What should they remember?
- What should they do?
If you cannot answer those questions, animation will not fix the story.
Give each scene one purpose
Label every planned scene with a verb: introduce, define, compare, prove, demonstrate, reassure, or ask. The verb keeps the script focused.
A scene that tries to define a problem, prove demand, explain a feature, and request a meeting will feel rushed. Split it. A scene that has no clear verb may not be needed.
Instead of saying “Market opportunity,” explain why the previous point leads to the size or timing of the opportunity.
Write for the ear
Spoken language needs shorter sentences and clearer signposts than a report. Read every paragraph aloud. If you run out of breath or lose the subject, rewrite it.
- Prefer familiar words over formal equivalents.
- Introduce acronyms before using them.
- Round numbers when exact precision is not the point.
- Use a pause before the important conclusion.
- Remove throat-clearing phrases and repeated setup.
Do not read the slide
If the screen says “Three onboarding bottlenecks,” the narration can name why they matter or guide attention to the first one. It does not need to announce the heading and then recite all three labels.
Sync the script and visual
Mark the phrase that should trigger each visual change. A number should appear when it is spoken, not several seconds earlier. A highlighted region should remain active long enough to be found and understood.
Keep important on-screen text visible after the narration moves on. Viewers read at different speeds, and captions also need space.
Edit for time, then record
Time a rough read before producing the voice. A natural delivery usually needs more space than a silent read. Cut repetition before increasing speed. Fast narration makes complex material harder to trust.
Finally, verify names, technical terms, and pronunciations. Add captions, then watch once without sound and once without looking away from the main frame. Both passes reveal different problems.
Cynthara Works will pair scene planning with narration so the words and visuals can be reviewed together. Join the waitlist to follow the build.