Runway Prompt Format: Scene-by-Scene Templates

Teams often fail in Runway because prompts are descriptive but not operational. A better approach is to use a scene-by-scene template derived from prompt from video workflows.

Core Runway prompt structure

Use this order:

1. subject and action
2. environment and time
3. camera motion
4. lens/depth clue
5. lighting and mood
6. temporal transition

Scene template you can copy

[Scene Goal]
Subject:
Environment:
Camera:
Lens/Depth:
Lighting:
Mood:
Transition:

Example

Scene Goal: Product hero reveal
Subject: A matte-black smartwatch rotates on a glass block
Environment: Minimal studio at dusk
Camera: Slow dolly-in from medium to close-up
Lens/Depth: 50mm feel, shallow depth
Lighting: Soft rim light with subtle blue fill
Mood: Premium, precise, quiet confidence
Transition: End on logo focus and fade to black

Optimization rules for Runway

• keep one dominant camera action per scene
• avoid mixing documentary and ad-style in one line
• keep transitions explicit between scenes
• keep negative prompt short and specific

FAQ

Should I write one long prompt or multiple scenes?

Multiple scenes with clear boundaries are easier to control.

Can I reuse one template for different products?

Yes. Keep the structure, swap subject and mood blocks.

Next reads

• Video to Prompt home
• Shot-language framework
• How to get prompt from video
• Team quality checklist