Why image-first usually wins
Image mode lets you validate subject, composition, and lighting quickly. Once those are stable, video only needs motion logic.
That reduces iteration cost, especially for branding, e-commerce, and character-driven work.
When video-first is valid
If the core requirement is movement quality rather than single-frame aesthetics, video-first can be efficient.
Even then, start with short motion descriptions and add visual complexity later to avoid compounded instability.
Hybrid workflows are most practical
A common high-performing pattern is image-first for style lock, video for motion validation, then back to image for final derivative assets.
This supports both campaign coherence and long-term asset reuse.