Last weekend, while waiting to board my flight from Ft. Lauderdale to Utah, I stumbled upon Delta's latest marketing campaign. Intrigued by the visuals, I decided to take on a creative challenge: replicating the campaign's output using artificial intelligence. In this article, I'll walk you through my workflow, present the results, and share valuable insights gained from this process.
Before diving into the process, let's take a look at the original Delta advertisements that inspired this project. These visuals provided the creative foundation for my AI-generated content, serving as a reference to compare the results. Below, you'll find images of the ads I encountered at the airport, which sparked my curiosity and led to this fascinating exploration of AI photography.
The Case for AI Photography:
- Cost Reduction: Traditional photoshoots can be pricey, involving photographer fees, multiple locations or sets, and hiring models.
- Customization: Stock photos often lack the specificity and flexibility needed for effective marketing campaigns.
- Efficiency: AI allows for faster exploration of different styles, quicker ideation, and reduced time to production.
A Streamlined Workflow:
- First, I designed a poster template in Figma, defining the colors, typography, and layout.
- Next, I used Midjourney to generate similar images. Currently, this process remains iterative. While it's easy to create content in Midjourney by giving creative control to the AI, producing content that adheres to your specific vision is more challenging. This is where effective prompt design emerges as a crucial skill.
- Finally, I upscaled and added my chosen images to the Figma templates for export.
- Cost Savings: AI-generated content can be produced at a fraction of the cost of organizing a traditional photoshoot.
- Rapid Ideation: Quickly explore hundreds of options to find the perfect fit.
- Creative Inspiration: AI models often present unexpected interpretations of your creative direction.
- Diversity: AI-generated content makes it easier to produce visuals that accurately represent your target audience.
- Framing: Limited control over subject placement within the frame.
- Text Rendering: Midjourney struggles with generating coherent text.
- Model Control: Unlike real models, AI-generated subjects lack the same level of fidelity in posing and expression.
- Low-Resolution Output: For large prints, Midjourney's output files may be insufficient, requiring the use of a third-party AI upscaler, like CodeFormer, which I used for this project.
In summary, AI photography offers significant advantages in cost reduction, customization, and efficiency. While there are still challenges to overcome, the creative potential and accessibility it provides make it an exciting frontier for the marketing industry. My experiment with Delta's marketing campaign is just one example of how businesses can harness AI to revolutionize their marketing strategies.