"Hi Samantha, I need new shoes for my trail race next month, can you get some delivered?" This could have been a question from Joaquin Phoenix to his AI Assistant, or lover, personified by Scarlett Johansson's voice in the Hollywood blockbuster H.E.R. 

 

It's also a request consumers will likely be able to make in the near future to the trove of AI tools being built on top of Large Language Models (LLMs).

 

The question marketers need to answer is: How can I make Samantha, vis-à-vis LLMs, recommend my brand?

 

The speed at which consumers have adopted GenAI tools such as ChatGPT or Gemini is proof that there is a need for a different kind of interface to discover knowledge, brands or products. AI-based results are progressively being embedded in day-to-day user activities like search (e.g. Google’s AI Experiments and Microsoft's Copilot) or shopping (e.g. Amazon’s Rufus or Instacart’s ChatGPT plugin). The range of tasks these LLM-based applications cover will only increase over time, driving the need to understand how brands are positioned and perceived by foundation models which include LLMs, presenting yet another challenge for CMOs.

 

The ability to measure campaign performance is at the heart of any marketer's role, so understanding the key metrics of success is paramount. With SEO now a long-established and integrated practice, GenAI is presenting a new challenge - and opportunity - to marketing teams and the brands they represent. And with it comes a new and potentially invaluable metric.

 

 

The function of a GenAI-driven Large Language Model (LLM) can be likened to a psychoanalysis of the web, in that it ingests all manner of online sources - product pages, reviews, blogs - to generate an accurate representation of what people think, good or bad. For instance, whether or not amateur runners perceive a certain sports brand as the most comfortable shoe to train in, or which face cream best suits a specific skin type. 

 

For marketers, who customarily invest hundreds of millions in surveys and panels in order to gauge consumer sentiment, measuring their AI Brand Awareness will be key in understanding what consumers really think and see on the web. 

 

And just like the early days of SEO, we need to assess, measure and track how these new algorithms position and perceive brands to understand how the prevalence of LLMs will alter marketing approaches:

It’s early days for model optimisation, but as an industry, we’re already making good progress in our quest to provide the tools needed to address these questions. 

 

 

The opportunity underlying the move towards LLMs is enormous. First and foremost, we need to focus our energies on trying to measure how much of the customer's experience is driven by GenAI, and then to understand how consumer perceptions are altered along this journey. Armed with this level of data, marketers can begin to devise approaches to impact how all the different models perceive a brand and its products and categories.

 

Just as ChatGPT allows user-level customisation far beyond a keyword search, so  GenAI-driven searches will likely open up the potential for marketers to leverage and impact the results delivered for user-level product and brand recommendations. In time, this will pave the way for a new era of ‘LLM optimisation’, to coin a second term.

 

 

Similar to the early days of SEO, tracking how large language models (LLMs) evolve over time and figuring out how to influence them will lead to a new category of services and tech tools.

 

How the major industry players will reconcile their traditional models with these new AI-driven opportunities will become clearer in the coming months. What is certain is that adtech vendors and marketers will need to shift their mindset in response to the rapidly evolving GenAI landscape. Measuring how these models perceive a brand and what drives their product recommendations will become an essential task for every marketing team.

 

As GenAI is progressively embedded across all consumer-facing applications, AI Brand Awareness might become the most important metric to track as a marketer.