AI brand agent, said Joanne Z. Tan, founder of AIXD.world, is key to brand visibility with AI search by LLMs SOM (share of model), brand content strategies.

How to Use an AI Brand Agent to Achieve AI search brand visibility, on LLM with “SOM (Share of Model)”

Joanne Z. Tan, AIXD.world, shows how to use an AI Brand Agent – in-house or 3rd party – to elevate brand visibility with AI search, on LLM with SOM (Share of Model) for companies. As people are turning to AI for search, and soon purchasing as well, brands must evolve from SEO (search engine optimization) to SOM strategies with problem solving.

To watch this as a 13-min video

To listen as a 13-min podcast

AI is already replacing traditional search. In a survey of 12,000 consumers, the Harvard Business Review (HBR) found that 58% had used AI to search for products and services in 2025, compared with 25% in 2023. The consumers using AI search tend to be younger, wealthier, and better educated than average.

Bigger changes are coming as agentic AI begins to change how people purchase products, not just find them. To adapt, brands need to learn how to use agentic AI, how to convince customers to use in-house “brand agents,” and finally how to convince third-party AI agents (from ChatGPT and others) to choose their brand’s products and services. 

In this article, we will examine strategies to maintain brand visibility in the age of AI.

What is “share of model” (SOM)?

In AI search, “share of model” (SOM) measures how often brands are mentioned in the results produced by different large language models (LLM), according to the Harvard Business Review. SOM tries to capture brand mention rates, average position, and brand perception. There is still no single, widely accepted method for measuring SOM, which can vary both by LLM and by the prompts used.

SOM is an evolution of “share of search” (SOS), which measures customer demand based on the number of queries for a particular brand. In the age of search engine dominance, SOS was a useful way to measure customer demand and tailor search engine optimization (SEO) efforts.

But there is a critical difference in brand visibility when comparing traditional search results and LLM (or AI) search. As the HBR authors write: “Failure to register on an LLM means a brand doesn’t appear at all before consumers. On ChatGPT, unlike Google, there is no ‘page two.’” (Emphasis added.)

How do brands become visible to LLMs?

The single best way to improve brand visibility in LLM results is to state a problem and solve it. According to the HBR: “LLMs are not optimizing for attention, they are optimizing for resolution. Identifying the ‘job to be done’ thus becomes the number one priority for brand leaders if they want to score big on SOM.”  There are two elements driving SOM – solving problems and demonstrating authority.

Solve specific problems.

Because LLMs are looking for “resolution,” rather than keywords, brands need to identify specific user needs, use cases, and pain points – and then offer specific solutions for each.

Instead of offering “the best running shoes,” for example, a brand is better served by offering “running shoes with a carbon fiber midsole for improved long distance training.” The second description offers a specific use case and “resolution.”

Establish authority.

LLMs also look for authority – signs that your brand has the expertise to answer questions and solve problems.

  1.   Show sources. One way to show authority is to include references to published research or other primary sources of information. A skincare brand that links to clinical studies is one example.
  2.   Product details. Brands can also show authority by providing detailed product information, lists of ingredients, product explanations, and more. The more your brand can help the LLM find “resolution,” the more brand visibility it will enjoy. 
  3.   User content. A third way to show authority is to highlight user generated content, comments, and reviews. A brand with an ecosystem of apps where users can interact is a natural way to boost comments and reviews to gain LLM attention, according to the HBR authors.

How can brands prepare for agentic AI?

In a 2026 Harvard Business Review article, “Preparing Your Brand for Agentic AI,” the authors argue that AI is changing not just how consumers find products and services, but how they buy them. 

The authors write, “Every major AI company is developing agents in anticipation of mainstream adoption.” As one example, OpenAI is partnering with payment platforms like PayPal and retailers like Walmart “to facilitate purchasing within ChatGPT.”

To maintain brand visibility and avoid becoming dependent on third party agents (which the authors call “consumer agents”), brands should consider developing in-house AI agents (which the authors call “brand agents”).

Three stages of agentic AI adoption

When should brands use an AI agent?

Brands should begin by asking under what circumstances their customers are willing to interact with an AI agent. The answer often depends on context. People are most likely to accept an AI agent for “low stakes” decisions like restocking household supplies.

The calculation changes when the stakes are higher (such as healthcare or financial decisions), involve personal connections (such as choosing gifts or personal items), or when human interaction is part of the buying experience (such as shopping for luxury goods). 

In these higher stakes situations, brands need to adjust, as we discuss below.

Convincing customers to use your agent.

The first challenge is to convince customers to use in-house brand agents, rather than general purpose consumer agents like ChatGPT, which may direct purchases to competing brands and products. Consumers tend to trust consumer agents as being “on their side,” but brand agents have advantages to leverage:

  1. Product and customer data. Brand agents have access to comprehensive product and customer data not available to consumer agents. One example is the beauty brand Sephora, which can make product recommendations from proprietary information on 140,000 different skin tones and 34 million user profiles, according to the HBR authors. Consumer agents can’t compete with that level of personalization.
  2. “Human in the loop” model. Unlike consumer agents, brand agents can keep a “human in the loop” to answer questions and solve problems when needed. Brand agents that can quickly bring a human into the conversation to add empathy and creativity have a natural advantage – particularly in “high stakes” transactions.
  3. Responsible AI practices. Two of the biggest consumer concerns about AI are privacy and auditability. Privacy means protecting customer data as a top priority. Auditability is the ability to track, verify, and explain the AI agent’s decisions. Brands that adopt and communicate responsible AI practices dramatically increase the adoption of brand agents, the HBR authors report.

Convincing other AI agents to choose your brand.

Since some customers will always prefer using consumer agents, the next challenge is to improve the odds those agents select your brand. 

  1.   Educating the AI. The first strategy to gain brand visibility is to “educate” LLMs and AI agents about your brand’s products and services. The HBR authors give the example of an alcohol brand that found incorrect product information in AI search results.

The brand began prompting the leading LLMs regularly and compiling their responses. It then began to update websites, advertising, and other content until the LLMs started to reflect correct product information and brand messaging. (We at 10 Plus Brand have been creating and updating SOM content for our clients to ensure their brands lead AI search. Contact us for more information.)

  1.   App integration. Another strategy is to integrate brand and consumer agents. One example is Instacart, which developed a ChatGPT plugin, as well as a ChatGPT-powered tool on its company website. When consumers ask for a recipe, Instacart can automatically add the ingredients to a shopping cart for delivery. Note that Instacart is integrated with ChatGPT and its strategy could be difficult to implement across multiple consumer agents.
  2.   Formatting and coding strategies. Another way to influence AI results – whether an LLM or a full AI agent – is through formatting and coding. The authors discuss two examples:  “Strategic text sequences” (STS). STS is “algorithmically generated” text that can be added to product descriptions, according to a Harvard Business School study. The STS text looks random, but it can influence LLM results – and may even override the directions in an AI prompt. STS is controversial and there are no defenses to it yet, the authors report.  The LLMS.TXT format. A more benign approach is the LLMS.TXT format, machine readable text that highlights the website content intended for LLMs and AI agents to review. The format makes it easy for AI to find the most accurate, authoritative, and up-to-date information. LLMS.TXT is an emerging format but it is beginning to gain acceptance.
  1. Pay-to-play. Advertising and promotion is not yet a standard feature of AI search but could become one, as it did in search engines. In the pay-to-play model, AI agents may be optimized to promote sponsored products and services. Brands will need strategies to adapt, the authors warn, while “ensuring transparency around paid promotions, in line with evolving global regulations.”

AI has already transformed search and may soon do the same for purchasing. To maintain brand visibility, organizations must update their strategies.

👉 contact Joanne’s team at AIXD.world to begin building your brand’s AI presence before your competitors do.

👉 Subscribe to our FREE Newsletter for more insights on AI agent for AI search, SOM — Share of Model, and AI experience design.

About the author,  Joanne Z. Tan, Brand Strategist, Thought Leadership Coach

Joanne Z. Tan is the Founder & CEO of 10 Plus Brand, Inc. Joanne is a globally recognized brand strategist, thought leadership coach, content & branding expert, and speaker. She helps founders, CEOs, executives, board members, leaders, entrepreneurs, and organizations decode their Brand DNA, elevate merely successful businesses to become powerful brands in the AI age. Joanne was trained in law and business, and had a liberal arts education from Brandeis University before earning a law degree. Her coaching emphasizes comprehensive strategies, business modeling, multidisciplinary thought leadership and high authority content creation, brand building, culture, GTM, user experience design, AI native brand architecture™, and AIXD™ (AI experience design). A former journalist, award-winning photographic artist, Joanne is also a poet, writer, and an avid wilderness backpacker.

© Joanne Z. Tan, 2026. All rights reserved.