AI retouching: How We Generated 720 Product Images and Saved 60% of the Budget

AI is becoming a massive game-changer in e-commerce, but it’s not just about writing product descriptions or search enhancements. It’s about solving logistical nightmares.

One of our recent clients, Best In Show Trolleys – an incredible US-based brand producing high-end equipment for dog shows – faced a problem familiar to many custom manufacturers.

They offer their products in 36 different colors. Because they manufacture on demand, they don’t stock every single variation. This created a major bottleneck:

1. Photoshoots were impossible: Photographing every product in every color would require enormous warehousing and time.
2. Manual editing was too expensive: Their products are complex, and it’s impossible to change the color uniformly for all product details. Expert retouchers estimated outrageous costs to manually recolor these images.
3. Poor UX: A simple color sample that was available on the website wasn’t enough. Customers struggled to visualize what a “Sparkle Purple” trolley actually looked like. 

The Solution: AI-Powered Image Generation

We knew that standard Photoshop work would kill the budget and timeline. Instead, we turned to a custom solution based on AI retouching.

We developed a workflow to automate the recoloring process while maintaining quality and preserving all product details. The goal wasn’t just to “tint” the image, but to replicate complex textures – matte finishes, metallic sheens, fabrics, and even sparkles.

How we did it

• The Input: We took a single high-quality source image and photos of the actual material swatches.
• The Process: We wrote a custom script to communicate with Gemini. We fed it the source image and the texture sample.
• The Output: The AI generated the product in the new color, preserving shadows, lighting, and texture.
• The Automation: The system automatically named the files to match the store’s SKU logic, allowing for instant assignment to the correct product variant.

The Results

By leveraging AI retouching instead of traditional photography or manual editing, we delivered a massive upgrade to the store’s User Experience in record time.

• 720 Images Created: Every variant now has an accurate product photo.
• 70% Time Saved: We completed the project in a fraction of the time estimated for manual retouching.
• 50-60% Budget Saved: The cost was significantly lower than the market rates for expert photo editing.

Most importantly, the customer can now clearly see exactly what they are buying, removing the guesswork and increasing conversion confidence.

Do you have a product catalog that is impossible to photograph?

If you are struggling with product variations and “imagination gaps” for your customers, the Omnycode team can help you implement a similar solution.