What makes a useful short video?

What you will learn

By the end, you will understand:

Choose media by communication job rather than trend
Plan accessible equivalents and responsive delivery
Secure rights and preserve truthful context

Visual explainer

See the idea clearly.

One idea passes through a prism into a still image for spatial detail, video for change over time and audio for voice and sound, with accessibility and rights surrounding every output.

Different media carry different information

Medium	Strength
Image	Composition, comparison, appearance, place, detail and a moment that can be inspected.
Video	Movement, sequence, demonstration, behavior, emotion and change over time.
Audio	Voice, conversation, sound, intimacy and screen-free listening.
Text	Precise reference, scanning, search, translation and flexible assistive access.

Choose the smallest medium that explains well

Do not make a five-minute video for a fact that one diagram explains. Do not use a still image when the essential lesson is a physical sequence. Production effort should follow communication need, not platform fashion.

Combining media can be powerful: a real photograph with annotations, a demonstration video with a transcript, or audio with a visual summary. Each layer should add meaning.

Image decisions

Clear subject and crop
Genuine relevance
Sufficient resolution at display size
Responsive source sizes
Text not baked into images unnecessarily
Alt text describes meaningful purpose
Decorative images have empty alt
Color is not the only signal
Rights and model/property permissions checked
AI-generated or edited context disclosed where needed

Video decisions

01
Define viewer task
02
Plan shots and evidence
03
Write spoken/visual script
04
Record readable sound and image
05
Edit for meaning
06
Add accurate captions/transcript
07
Review claims and rights
08
Export efficient formats
09
Test mobile and reduced-data context

Audio decisions

Clean intelligible speech matters more than expensive equipment. Control echo, noise and microphone distance; provide structure and verbal descriptions when a listener cannot see what speakers reference.

A transcript improves reference and access, but should identify speakers and meaningful sounds. Music can set mood but requires appropriate rights and must not overpower speech.

Accessibility changes production, not only publishing

Need	Approach
Image alternative	Purposeful alt text or adjacent explanation; complex diagrams need a fuller text equivalent.
Captions	Accurate synchronized dialogue and meaningful sound for prerecorded video.
Audio description	Describe essential visual information not conveyed in the main audio where required.
Transcript	Provide text for audio/video content with speaker and relevant sound context.
Motion safety	Avoid flashing and give control over moving or autoplaying content.

Rights attach to ingredients

Photographs, footage, music, sound effects, performances, locations, brands and people may involve different rights. “Royalty-free” does not mean no licence conditions; platform music permissions may be limited to that platform or use.

Keep asset source, licence, creator, consent and edit records. Do not assume an AI tool makes training inputs or outputs free of copyright, publicity, privacy or trademark risk.

Protect truthful context

Cropping, editing, synthetic voices and generated scenes can change meaning. Do not present staged, reconstructed or synthetic material as documentary evidence. Disclose material alterations when omission could mislead.

For testimonials, the speaker’s experience and brand relationship must be genuine and clearly disclosed where material.

Real-world example

Example: choosing media for a product repair

Example

A repair company uses one annotated image to identify the reset button, a 40-second captioned video to show the reset sequence and a text checklist for safety conditions. A podcast would be inefficient because the task depends on seeing the device.

Try this

Match idea to medium

Choose one planned topic. Write what must be seen, heard, read and demonstrated over time. Remove every medium that does not add useful information, then list the accessibility and rights work for those that remain.

Common questions

Questions beginners ask.

When should I use video instead of an image?

When movement, sequence, behavior or change over time is essential to understanding.

Does every image need alt text?

Meaningful images need an appropriate alternative; purely decorative images should normally use empty alt text so assistive technology can ignore them.

Are automatic captions enough?

They are a starting point but should be reviewed for names, technical terms, punctuation, speakers and meaningful sounds.

What is audio description?

Spoken description of essential visual information that the main audio does not communicate.

Does royalty-free mean free to use anywhere?

No. It usually means a licence avoids per-use royalties, but conditions, scope and attribution may still apply.

Can I use music from a social platform elsewhere?

Not automatically. Platform licences and commercial-music rules may limit where and how it can be used.

Should text be placed inside images?

Avoid it for essential information when real text can be used; real text is more responsive, selectable, translatable and accessible.

Do AI-generated images need review?

Yes—for accuracy, bias, harmful representation, rights, brand fit and whether disclosure or provenance metadata is appropriate.

Assessment

Check what you understood.

5 questions · instant explanations

1. Which medium best shows a physical four-step repair?

2. What should happen to automatic captions?

3. What does “royalty-free” mean?

4. How should a decorative image usually be exposed to a screen reader?

5. True or false: an AI-generated photograph-like scene can always be presented as documentary evidence without context.

Sources

How do images, video and audio communicate differently?