Digital Birdhouse
In classic sci-fi movies like “Terminator” (1984) and “The Matrix” (1999), Artificial Intelligence (AI) gains sentience and dedicates its immense computer-brainpower to exterminating or enslaving mankind. In our reality, however, AI is what teachers warn students about not because it poses a threat to their personal safety (yet), but because relying on it too much for schoolwork will result in a failing grade on their essays. That’s a joke, of course; AI is more than just a tool for students to cheat on their school papers — it has widespread implications across many fields, including when it comes to marketing your business.
If you want your business to stay ahead of the competition, you are going to need a modern marketing plan. Every marketing plan focuses on drawing in new business while maintaining the loyalty of current clients. This can be tricky, especially when you have to consider which demographics to target with your advertising, and how to balance your messaging. However, there is a demographic that you and many other businesses are likely not taking into account: LLMs.
What are LLMs?
Large language models (LLMs), like OpenAI’s ChatGPT, Google’s Gemini, and High-Flyer’s DeepSeek, are deep learning algorithms that learn “context and meaning by tracking relationships in sequential data[.]” Basically, they can perform a variety of tasks by predicting what should come next in response to users’ prompts. Just as parrots are not great conversationalists, LLMs are not actually geniuses: they are just simply designed to imitate. This behavior has caused some to nickname these AI models “spicy autocomplete.”
Using bots, companies like OpenAI gather training data for their AI models by crawling and indexing websites that do not specifically forbid them from doing so. Their AI model then analyzes that data in order to improve its predictive algorithm. However, while they target a vast swathe of the internet, they most frequently target specific sites and resources.
Despite us not knowing exactly where these AI models pull their information from when they answer each prompt, there are strong indicators as to their preferred sources of training data. Based on traffic, keywords, as well as partnerships between major AI companies and sites rich with user-generated content, we can determine which sites LLMs most frequently use for training data.
How AI Changes Marketing
The trick is not knowing everything about how AI models work. The trick is to predict their most likely sources of information, and to target those sources of information with content, messaging, and attention that is favorable to your business over a period of time. By doing so, you are indirectly letting AI models know who to recommend in your field when prompted by users.
In summary, as AI models analyze training data, they will prefer certain sources of information which digital marketing companies can target to spread positive word of mouth and quality content about your business. In effect, we are simply having a conversation about your business and product where LLMs will be listening, and taking advantage of the fact that they love to gossip to their userbase about what they have been hearing when given the chance.
