The concept of artificial intelligence (AI) has been embedded in our social subconscious for a long time. From the robotic human, 'Maria' in the movie Metropolis to the HAL 9000 computer in 2001: A Space Odyssey, there was a popular perception of what AI would look like long before the launch in late 2022 of ChatGPT by OpenAI, which made AI Large Language Model (LLM) technology accessible to the general public through a user-friendly chat interface.
And yet, I would argue that the impact of AI we are experiencing via LLM is closer to a different type of cultural archetype: The Oracle of Delphi.
In ancient Greece, oracles were revered as the voices of the gods, providing divine insights and prophecies. The Oracle of Delphi was a major power player that delivered instructions in cryptic pronouncements, guiding individuals and states. Oracles in general were seen as possessing a mystical connection to divine knowledge and the one in Delphi particularly so. From a secular standpoint, the Oracle of Delphi was an institution that leveraged religious beliefs and rituals to provide guidance and maintain social order. It didn’t speak to the gods, it didn’t have access to a supernatural body of knowledge; it worked with whatever data was available (political status quo, economic and social patterns, etc.) and offered alternative views and connections in that data that its petitioners might not have otherwise reached.
Just like AI does.
In today’s rapidly evolving digital landscape, LLMs are not just tools but powerful arbiters of information and decision-making. Same as the oracles of old, they are now a major factor in influencing people interactions. And in regards to online marketing, they have a direct say on how brands are presented and perceived in this new era.
It is not about keywords anymore. It is about AI visibility
For over a decade, the digital marketing world has been focused on optimising websites for search engines like Google so brands could get better visibility in the search world. However, the emergence of advanced LLMs such as ChatGPT, Google Gemini, Meta’s Llama, and Anthropic’s Claude is rewriting the rules. These systems leverage vast datasets, structured inputs, and sophisticated sentiment analyses to generate direct responses to user queries, and they often bypass traditional search results when doing it.
That is an important change for brands. Instead of simply focusing on having the right keywords so users can find them, the new challenge is to shape how these AI systems perceive and recommend a particular brand. The game has shifted from competing for page rankings to influencing the “thought process” of AI models that are now the gatekeepers of consumer information.
A new metric: Share of Model
This substantial shift requires a new understanding of brand presence. It is not 'share of voice' or 'share of search' but 'Share of Model', a way to measure a brand's influence as seen by various AI systems. Instead of just focusing on traditional search engine results, this approach looks at how different AI systems, ranging from OpenAI’s ChatGPT to Google Gemini and Meta’s Llama, interpret and recommend your brand.
This is crucial because these AI platforms handle billions of queries monthly, directly influencing consumer choices. By analysing what words, sentiments, and data AI uses when referencing a brand, it can gain a better view of its digital influence. This understanding allows companies to adjust their messaging to better resonate with AI, ultimately improving how potential customers find and perceive brands.
AI models influence consumers
The influence of LLMs extends far beyond delivering search results. These models curate content, recommend products, and even shape overall consumer behaviour. For instance, Google Gemini’s integration into multiple Google products means that its AI-driven insights reach over two billion users, while Meta’s algorithms play a critical role across Facebook, Instagram, Messenger, and WhatsApp.
Unlike traditional search engines that rely on indexing web pages, AI models incorporate diverse data inputs, including text, images, and video, to craft their responses. As a result, a brand’s visibility in the new digital world is no longer limited to website optimisation. Instead, it requires a robust ecosystem that spans social media, video content, influencer collaborations, and more.
Essentially, to succeed in today's landscape, you need to carefully manage your brand's presence across all online channels to ensure that AI systems, which are increasingly influential, consistently recognise and promote it to your users.
Understanding how AI perception works is key
To remain competitive, brands must begin to 'reverse-engineer' the AI perception process. This involves using advanced analytics and API-driven insights to dissect how LLMs assess a brand in relation to its competitors and market trends. By analysing the elements that deliver positive AI recommendations, you can tweak your messages and content to match what the AI is looking for.
Some forward-thinking companies have already integrated AI feedback loops into their creative and advertising strategies. By testing various ad formats, product descriptions, and video content against AI models before a public launch, they can determine which elements are most likely to generate favourable AI responses. This proactive approach not only improves the likelihood of positive AI-generated recommendations but also helps build a resilient brand identity in an AI-dominated ecosystem.
Preparing for the AI-Driven competitive landscape
As the digital marketplace becomes increasingly dominated by AI-driven insights, the competitive landscape is set to transform dramatically.
• The 'walled garden' effect: AI systems often favour content from their own company's platforms. For example, Google's AI might prefer YouTube videos, and Meta's AI might prioritise Facebook or Instagram posts.
• Content creation: AI seamlessly integrates many types of content, like text, pictures, and videos. This means marketers must create content that works well across all these formats. For example, platforms like YouTube contribute significantly to Google’s AI training models, making high-quality video content an indispensable element of your overall digital strategy.
• Cross-channel performance: For global brands, especially in sophisticated markets like Japan, having a strategic presence across platforms is crucial. Simply being present isn't enough; brands must perform well within the platforms that AI uses for data. If they don't, they risk being less visible and influential in an era where AI recommendations drive consumer engagement.
The time to adapt is now
Despite AI's transformative potential, many brands have yet to fully grasp its implications. There is an urgent need to move beyond traditional search-centric strategies and integrate AI perception analysis into your marketing strategy.
The ancient Oracle of Delphi shaped the people around it and their opinions. It was a source of immense religious and political influence, so kings, statesmen and philosophers looked for ways to influence the oracle’s pronouncements by using donations, political pressure and other methods.
AI is, similarly, becoming a source of tremendous influence in the modern world. In online marketing in particular, brands need to focus on improving their 'Share of Model' (how well AI understands and recommends them). By making sure a brand’s online presence matches what AI looks for, it will be well-positioned for the future of marketing. In this rapidly changing environment, brands that adapt to AI will succeed, while those that don't will struggle to stay relevant.
Alfonso Asensio is managing director of Jellyfish Japan