Generative AI in marketing: how NZ and UK businesses can use AI imagery, video and design

For years, great visuals were one of the biggest gaps between small businesses and big brands. A product shoot, a location, models, lighting and retouching could cost more than a whole month's marketing budget. Generative AI has changed that. With the right brief and an experienced eye, you can create imagery and video that fits your brand, for a fraction of the time and cost.

This guide explains what generative design means in marketing, where it works well, where it doesn't, and what businesses in New Zealand and the UK should know before using it. If you'd like us to create visuals for you, see our AI imagery, video and generative design service.

What is generative design in marketing?

In marketing, generative design means using AI models to create visual content from a brief: text prompts, reference images, brand colours, product photos and creative direction. The output can be still images, short videos, 3D scenes or design variations.

The term has an older meaning too. In engineering and architecture, generative design describes software that explores many design options based on goals and constraints, such as weight, strength or cost. The idea is the same: you set the direction and the rules, and the software generates options for a person to choose and refine.

What generative AI can do for your marketing

1. Product imagery and lifestyle scenes

Place a product in a styled scene, a seasonal setting or a lifestyle context without a physical shoot. This works especially well for ecommerce, jewellery, homewares, fashion accessories and packaged goods.

2. Campaign and social visuals

Create a consistent set of on-brand images for ads, social posts, email and landing pages, and produce variations quickly to test which ones perform best.

3. Concept and pre-launch visuals

Show customers or investors something that doesn't exist yet: a building, an interior, a product range or an event. We've created AI visuals of Auckland high-rise architecture for an architectural firm, so people could see concepts before they were built.

4. AI video and storytelling

Short product videos, brand stories and social clips can now be generated and edited together without a film crew. For Alcyone Bespoke Jewellery, a London jewellery brand, we produced AI product videos for its collections, combining generative video with traditional editing and storytelling.

5. 3D scenes and immersive visuals

Real-time 3D tools such as Unreal Engine, combined with AI, can create detailed environments for campaigns, websites and presentations, like our "Auckland 2053" vision of a future city skyline.

6. Brand exploration

Explore visual directions, moods and styles early in a branding project, so decisions are made with real examples in front of you rather than abstract descriptions.

Where generative AI still falls short

• Accuracy of real products. AI can subtly change shapes, logos, text and details. Anything showing a real product needs careful checking and often retouching.
• Brand consistency. Without clear direction, AI output drifts in style from one image to the next. Consistency comes from a defined brief, references and an experienced eye.
• Hands, faces, text and small details. These have improved a lot, but still need close inspection before anything is published.
• The generic look. Unedited AI images can look obviously artificial or similar to everyone else's. Good art direction is what makes them feel like your brand.
• Real people and real places. Be careful generating images of identifiable people, and don't present AI scenes as photos of real events or locations when that would mislead.

Generative AI in New Zealand and the UK: what to know

The rules are still developing, but a few principles apply in both countries:

In practice, that means: keep product visuals true to the real product, check the terms of the AI tools you use, avoid copying another brand's style or protected work, and be thoughtful about disclosure. This isn't legal advice, so for anything high-stakes, check with your lawyer.

How we create AI imagery and video

Generative design works best when it's led by people who understand images and stories. Our process is built on years of image-making and video storytelling, using the best technology and tools for realism:

1. Brief. We agree the goal, audience, channels, brand style and what must stay accurate, such as product details.
2. References and direction. We gather references, product photos and brand assets, and set the look, lighting and mood.
3. Generate and select. We create many options and choose the strongest, rather than accepting the first result.
4. Refine and retouch. We correct details, match colours to your brand and fix anything that looks artificial.
5. Check and deliver. We check accuracy, consistency and suitability for each channel, then deliver files sized for web, social, ads and print.

AI helps us inform, review, automate and execute, with human-centred design and strategy directing every step.

How to get started

1. Pick one use. Start where visuals are holding you back, such as product pages, a campaign or social content.
2. Gather what you have. Product photos, brand colours, fonts and examples of visuals you like.
3. Set the rules. Decide what must be accurate, what's off-limits and how you'll handle disclosure.
4. Test and measure. Compare AI visuals with your current ones on click-through rate, engagement and conversions.
5. Build a library. Once a style works, reuse the same direction to keep everything consistent.