Producing a creative winner is not an easy job. There is this tactic when you create multiple different creative concepts - static images. You test them against each other. You take the winner and make a video out of that creative concept.
With the rise of AI, this can be even more effective. For two reasons:
You can build different static concepts super fast and move to video production
You can be inspired by what others made and iterate until you find a winner
I will guide you through some best practices, creative ideation process and tools I use and show you some creative iterations with real data. Let’s dive in:
Prompts
The Art is in the prompt.
Clear Direct Prompt = clear direct output.
Ambiguous Prompt = Ambiguous output!
How about the prompt structure?
Prompt Example 1:
A genuine, real-life moment of a unicorn with big horn on a battlefield, seen from behind. He stands at the edge of a battlefield, perhaps a little windswept, with his horn is shining. The battlefield isn't pristine; there are signs of other soldiers – a few footprints, a distant explosions, maybe a forgotten bazooka and dead bodies. The sky might have a few clouds, hinting at a day that's warm but not overly sunny. Medium: Photography. Style: Authentic, real-life. Lighting: Natural daylight, slightly overcast with diffused sunlight. Composition: Shot with a Nikon D850, using a 50mm lens, Resolution 45.7 megapixels, ISO sensitivity: 200, Shutter speed 1/250 second --v 5.1 --style raw --s 750.
Prompt Example 2:
realistic mystery horrorcore honeycore unicorn blue green pink galaxy volumetric lighting --v 5.2
AI Creative Iteration
So many different views, angles, places, backgrounds, and ideas that you can use to find a winner quickly. Let’s continue with the iteration process: