How Coca-Cola Is Using Generative AI to Rethink Global Creative Production

How do you make thousands of creative decisions in markets around the world, and avoid losing the brand you know and love? For Coca-Cola, this is becoming a more and more difficult task, as consumers continue to demand more relevant, personalized content via digital channels.

Generative AI could help here, but while creating more content is one option, having the bandwidth to churn out content without diluting the brand is the real challenge.

The brand’s work with AI, from its Project Fizzion design system to localized campaigns and product innovations, provides a clearer picture of how marketing teams can rethink content workflows, balance global consistency and local relevance, and put human creativity front and center.

The Shift from Experimental Campaigns to Core Workflow

Experiments conducted by Coca-Cola during its early explorations of generative AI revealed the potential of this technology. However, the challenge was for the technology to become an integral part of the way the company will produce and distribute marketing content. It is important since large corporations must stay relevant on multiple platforms and provide reasons for consumers to buy their product beyond convenience and affordability.

The company’s José vs. Mourinho series for the FIFA World Cup 2026 offers a recent example. Based on an AI-driven digital twin of José Mourinho, the video series includes two avatars of the football manager, each holding a different perspective on the game of football. The series is not dependent on one advertisement but provides scope for continuous, social first content.

The strategic value lies in the repeatable format. Once a creative concept works, AI can help teams explore more variations and respond to audience interests across digital channels. In any case, such a marketing effort does not provide evidence of AI having become a foundation stone of the whole marketing effort of Coca-Cola. This is the way the company is exploring the potential of artificial intelligence for creating content fast and relevant to the audience.

Also Read: How Patagonia Built a $1.5B Brand on a Sustainability-First Marketing Model

Scaling the Unscalable by Transforming Creative Production

Conventional creative production is slow because the creation of every new format, audience, or campaign iteration involves design, refinement, and approval processes. A team can have a great campaign idea; however, the conversion of that idea into various assets to fit the needs of different formats and regions can delay the process of its execution. Generative AI brings about a different dynamic since it allows teams to explore visual ideas, experiment with backgrounds, and build on top of existing designs without creating them from scratch every time.

Coca-Cola’s Project Fizzion takes this idea further. Originally developed as an intelligence design system, its StyleID tool converts creative designers’ decisions into intelligent design rules, enabling creative teams to create hundreds of localized designs while retaining consistency across the brand. According to Coca-Cola’s press release, the system was in the pilot phase when the company launched it.

What matters is how creative teams can spend their time. Rather than doing all the work themselves, designers can define visual rules for the system to apply them to various designs. Creative teams can therefore direct their attention to the more complex parts concepts and messages as well as variations requiring careful consideration. Such approach would also facilitate A/B testing through the generation of different versions of a marketing campaign; however, it is not clear from the press release that Coca-Cola uses Fizzion to generate hundreds of test variations.

The main indicator of success is going to be how improved productivity contributes to improved marketing decision-making.

Hyper-Localization Across Global Markets

If the brand operates in several regions, then localization is not limited to translation of the campaign into another language; imagery, culture, seasons, and consumer behavior will be different from region to region as well. As Coca-Cola works with bottling companies and regional marketing departments that are familiar with their audience, the company has to take all these factors into consideration. With generative AI technology, adaptation and generation of various versions of the campaign would become easier without the need to recreate all the assets from scratch. However, local marketing teams still have to review whether adaptation fits the target audience.

Chinese New Year Campaign by Coca-Cola can serve as an example of how the technology could be used for generating campaigns. Drawing inspiration from traditional Chinese paper art, the campaign offered to create digital portraits of one’s friends and relatives using connected packaging during the Chinese New Year holiday that celebrates family reunions.

This is not about creating greater amounts of localized content. It’s about linking a certain idea to a particular cultural environment in which the consumer can join in on this idea. This is important because, regardless of the accuracy of the translation, a campaign may lack that element of being relevant to the culture of the place.

For Coca-Cola, it means an opportunity to explore these kinds of experiences through generative AI, but it still takes a cultural understanding to make it relevant. There is no information about any performance data specifically related to the AI technology used by the company during this particular campaign.

Product Ideation Through the Coca-Cola Y3000 Case Study

Generative AI can influence more than how a product is marketed. It can also help shape the ideas behind the product itself. Coca-Cola tried this through Y3000 Zero Sugar, which is a limited edition drink released in 2023 through the platform of Coca-Cola Creations. It is a product that was co-created with human intelligence and AI in order to integrate consumer perspective into the development of a taste of the future.

Consumer perspectives from across the globe along with the use of AI insights inspired the flavor of this product. AI was also used to give this product a futuristic design and to shape the digital consumer experience. This has enabled a combination of product development and marketing with the same idea underlying the taste and experience of the consumer.

The difference between the inspiration provided by the AI and the use of AI to make decisions should be clear. The public statement of Coca-Cola does not state that the company scrapped massive amounts of social media conversation to figure out the flavor profile. The confirmed approach was a combination of global fan perspectives and AI-derived insights.

Y3000 shows one way Coca-Cola’s generative AI efforts can extend beyond content production. AI can help teams explore possibilities and translate an abstract idea into a product experience. Human judgment remains essential, particularly when those possibilities must become something consumers will actually want to buy.

Why Brand Strategy Needs People

Creative production using generative AI is not a matter of letting a machine make decisions regarding the branding strategy. While AI can help speed up the creative process by providing options for variations, adapting assets, and processing information, it cannot be relied upon to tell if the campaign works well for a specific audience, delivers the right emotions, or matches the overall vision for the brand in the future.

A great example of how Coca-Cola leverages generative AI is the reimagined Holidays Are Coming campaign. Although the company has used generative AI to develop two new versions of its holiday campaign, the fact that the global generative AI leader at Coca-Cola highlights that it was a human-driven decision-making and creative process with the use of generative AI in execution.

It is a perfect demonstration of the human-in-the-loop principle: AI helps expand creativity and speed up processes, whereas people determine boundaries and make final decisions about which of the options should reach the target audience. It is a useful distinction for any enterprise marketing team considering similar tools.

For Coca-Cola to fully realize its long-term benefit from generative AI, there is need for a combination of rapidity in production alongside good judgment. AI can assist in the process, but it is not going to take responsibility for the brand’s values.

Conclusion

Generative AI is influencing the way in which Coca-Cola is producing its creative content, from the creation of new campaigns to new product development and even locally relevant experiences. But soon, as more businesses acquire the same technology, the ability to distinguish oneself through those technologies will become increasingly difficult. What is going to make the difference is going to be the application of such technologies, along with the consumer insight and creative input behind the scenes. Coca-Cola’s exploration of generative AI indicates that the importance of production speed is something that cannot be ignored but is not everything by itself.