Your most important customer may be AI

Imagine you run a meal prep company that teaches people how to make simple and delicious food. When someone asks ChatGPT for a recommendation for meal prep companies, yours is described as complicated and confusing. Why? Because the AI saw that in one of your ads there were chopped chives on the top of a bowl of food, and it determined that nobody is going to want to spend time chopping up chives.

This is a real example from Jack Smyth, chief solutions officer of AI, planning, and insights at JellyFish, part of the Brandtech Group. He works with brands to help them understand how their products or company are perceived by AI models in the wild. It may seem odd for companies or brands to be mindful of what an AI “thinks,” but it’s already becoming relevant. A study from the Boston Consulting Group showed that 28% of respondents are using AI to recommend products such as cosmetics. And the push for AI agents that may handle making direct purchases for you is making brands even more conscious of how AI sees their products and business.

The end results may be a supercharged version of search engine optimization (SEO) where making sure that you’re positively perceived by a large language model might become one of the most important things a brand can do.

Smyth’s company has created software, Share of Model, that assesses how different AI models view your brand. Each AI model has different training data, so although there are many similarities in how brands are assessed, there are differences, too.

For example, Meta’s Llama model may perceive your brand as exciting and reliable, whereas OpenAI’s ChatGPT may view it as exciting but not necessarily reliable. Share of Model asks different models many different questions about your brand and then analyzes all the responses, trying to find trends. “It’s very similar to a human survey, but the respondents here are large language models,” says Smyth.

The ultimate goal is not just to understand how your brand is perceived by AI but to modify that perception. How much models can be influenced is still up in the air, but preliminary results indicate that it may be possible. Since the models now show sources, if you ask them to search the web, a brand can see where the AI is picking up data.

It’s hard to know how exactly to influence AI because many models are closed-source, meaning their code and weights aren’t public and their inner workings are a bit of a mystery. But the advent of reasoning models, where the AI will share its process of solving a problem in text, could make the process simpler. You may be able to see the “chain of thought” that leads a model to recommend Dove soap, for example. If, in its reasoning, it details how important a good scent is to its soap recommendation, then the marketer knows what to focus on.

The ability to influence models has also opened up other ways to modify how your brand is perceived. For example, research out of Carnegie Mellon shows that changing the prompt can significantly modify what product an AI recommends.  Change from recommending a specific brand could range from 0% of the time to recommending it 100% of the time. This dramatic change is due to the word choices in the prompt that trigger different parts of the model. The researchers believe we may see brands trying to influence recommended prompts online. “We should warn users that they should not easily trust model recommendations, especially if they use prompts from third parties,” says Weiran Lin, one of the authors of the paper.

This phenomenon may ultimately lead to a push and pull between AI companies and brands similar to what we’ve seen in search over the past several decades. “It’s always a cat-and-mouse game,” says Smyth. “Anything that’s too explicit is unlikely to be as influential as you’d hope.”

Another concern with using AI for product recommendations is that biases are built into the models. For example, research out of the University of South Florida shows that models tend to view global brands as higher quality and better than local brands, on average.

AI can also serve as a focus group for brands. Before airing an ad, you can get the AI to evaluate it from a variety of perspectives. “You can specify the audience for your ad,” says Smyth.

Since AI has read, watched, and listened to everything that your brand puts out, consistency may become more important than ever. “Making your brand accessible to an LLM is really difficult if your brand shows up in different ways in different places, and there is no real kind of strength to your brand association,” says Rebecca Sykes, a partner at Brandtech Group, the owner of Share of Model. “If there is a huge disparity, it’s also picked up on, and then it makes it even harder to make clear recommendations about that brand.”

Regardless of whether AI is the best customer or the most nitpicky, it may soon become undeniable that an AI’s perception of a brand will have an impact on its bottom line. “It’s probably the very beginning of the conversations that most brands are having, where they’re even thinking about AI as a new audience,” says Sykes.

show less