Why is it important to consider your company’s placement in LLM search results?

LLM search is taking market share from traditional search platforms – LLMs are gaining traction in the discovery space, indicating a significant shift in how users query and learn about new products online. Google’s share of search has slipped below 90% for the first time, and daily LLM usage exceeds that of social media platforms like LinkedIn by orders of magnitude.

Search Engine Data from StatCounter

Using AI search is correlated with behavioral shifts that change how customers make queries – users are more likely to make iterative, conversational queries to AI tools like Claude or ChatGPT when they’re at the very top of the funnel. Once they’ve used AI search to get a better idea of the market, users switch back to traditional search to dive deeper.

The AI search playbook differs from the traditional internet search playbook – the world’s best traditional SEO strategy won’t necessarily help you show up for AI searches. While web search was focused on the content on your website and how many links you have, AI looks at how much influence you have and what other people say about your content/brand.

How does Google search work differently than AI search?

Google searches are designed to drive traffic, whereas LLM searches present information that is pulled from multiple sources – companies have more direct control over the factors that influence Google search results (e.g., owned website content and website structure). If you aren’t showing up on Google search, you can optimize the owned factors that improve SEO performance. If you aren’t showing up through LLM search, the solution is more complex.

AI search relies on reputation and external validation – LLMs prioritize customer review sites, industry analysts such as Forrester and Gartner, and other verified third-party sources over the content a brand writes about itself. AI search optimization and PR are beginning to intersect due to the importance of consistent external validation.

The accuracy of AI search results depends on how a query is phrased – AI SEO remains less reliable because LLMs try to match patterns in tokens (data) out in the web to patterns in the language you use in your query, which can lead to hallucinations or inconsistent answers. For example, if you ask, “Is JP Morgan a central or state-owned bank?” an LLM might say that JP Morgan is a central bank, even though it is neither.

How Google vs. AI Search Results Differ
Search Type	Google	AI
Main factors that determine search results	Owned website content – site content should build authority and use industry terms to demonstrate relevance   Rankings – websites with higher traffic for relevant searches are more likely to show up	Amalgamation of the larger web ecosystem – data accumulated from multiple sources determines relevance based on reputation/ “what people say about you”
Brand adaptability	High – changing website content adjusts search results	Low – search results depend on many external factors you can’t control
Prioritization framework	Link graph	Social graph and link graph
Results format	Ranked results allow websites to include their own language	Short list of top results includes product/ company names

Optimizing AI SEO

How should you adjust your approach to website SEO to improve AI search performance?

Clearly explain features and benefits – use straightforward and extensive language to create strong token associations that LLMs can pick up on. For example, to enable their water-resistant jackets to better compete with waterproof competitors, Canada Goose created website content explaining why water-resistant jackets can meet the needs of customers who think they need waterproof jackets.

Use case studies to signal your target audience – case studies are an opportunity to describe the segment of the market that gets the greatest value from your product. AI search tools use statements such as “How Product X drove 30% growth for Customer Demographic Y” to determine who to present your product to and how to position it.

Leverage press releases and blog posts to support updates – AI search tools are more likely to recognize important updates if they are called out in a dedicated press release or cited in a blog post. Think of these as channels as a way to inform LLMs about your brand’s evolution.

Adopt token recognition techniques used in the “early days” of Google SEO optimization – don’t be afraid to describe your company in plain, repetitive language. While the tone of voice of your entire website shouldn’t change, this rudimentary approach helps LLMs identify patterns linking your product to your target demographic.

How should you adjust your approach to offsite owned content to improve AI search performance?

Use consistent messaging to direct the customer narrative – align the messaging and key talking points you use across all channels including your website, social profiles, review sites, and anywhere else you have publicly visible profiles. If what you stand for is obvious to customers, they will echo those sentiments in their own comments.

How should you adjust your approach to managing external voices to improve AI search performance?

Think of PR as an SEO tool – PR is still a valuable lever for increasing overall awareness, but its overwhelming implications for AI search may justify a larger monetary and resource investment. New products and companies struggle to appear in LLM search results without the support of a PR campaign.

Leverage strategic PR and industry relationships to secure important external coverage – LLMs value the opinions of industry analysts, respected publications, and influential reviewers in your space. If you can’t cultivate relationships with Forrest, Gartner, or other industry publications yourself, hire a PR partner that can.

Activate your community – companies that excel at evangelizing their communities have a significant advantage with AI search. Encourage your customers, advisors, and advocates to actively participate in conversations about your brand on LinkedIn, Reddit, G2 (formerly G2 Crowd), and other review sites to generate the volume and authentic content that LLMs value.

Stay up-to-date on how LLMs reference authoritative sites – sources like The New York Times block LLMs from using their content, and other authoritative sites might seek to protect their work from being accessed. AI tools might also consider sites such as Reddit to be authoritative, even if their content is less accurate than a company’s own website. 

How should you structure content to optimize for AI search?

The ideal content format depends on the LLM – LLMs are constantly changing how they prioritize and rely on different formats of content. Right now, Gemini uses YouTube to inform most of its responses, while Claude and ChatGPT value the written word over other media.

Prioritize content based on customer preference – any content you create for AI search purposes won’t only be consumed through the lens of a response from an AI tool. Look at your ideal customer’s media diet and start there. For example, if your ICP conducts most of their product research through YouTube, create video content first.

Which AI search-specific SEO challenges should companies be aware of?

LLM search makes change more difficult – Google’s search algorithm makes it relatively easy for new companies and products to get discovered. In contrast, AI search results are influenced by mention volume and how brands have been discussed in the past, which creates a hurdle for companies that are trying to pivot or break into a new space. Examples of things AI makes more difficult include:

• New company launches – it’s harder for new businesses to compete with the years of mentions and industry validation of more established competitors without significant PR investment.  
• New product launches – even positive reputations that relate to 1 product can impede the progress of a new product launch because LLMs associate companies with the products that first generated their success.
• The “Re’s” (rebranding, renaming, repositioning) – because LLMs prioritize volume, changing your name risks isolating your company from the successful momentum it’s garnered in the past. Effectively repositioning or rebranding is more difficult because you must overcome significant past associations/ mentions of your previous brand identity. LLMs are also slow to pick up on brands that move upmarket.

AI search makes it harder to compete with companies that have more reviews – competitors with significantly more reviews than you (e.g., 1,000 reviews vs. your 300) have a significant advantage because LLMs use review quantity as a measure of external validation. It’s difficult for new companies or companies with smaller customer bases to catch up to the massive quantities of reviews that more established competitors have generated over time.

How are user search behaviors changing with AI search?

Multi-search behavior is becoming the norm – users are more likely to ask conversational, iterative, and specific questions when querying AI search tools. Brands need to think more holistically about how to appear for a wider range of queries and provide LLMs with the “ammunition” to contextualize their result for more tailored searches.

Toggling between LLMs is becoming more common – Google has dominated search for a long time, which enabled marketers to optimize for a single framework. With AI search, however, brands should avoid optimizing for a single platform and strive for consistent representation across the entire AI search ecosystem.

Expect user behavior to change again as LLMs monetize – unlike traditional search, each AI query incurs a marginal cost to the provider. Future monetization strategies could impact how frequently or extensively users engage with these tools. 

Note: due to the prominence of multi-search behavior, daily active users can provide a more useful perspective on AI search than number of searches.

How do you measure AI search performance? What are the most important metrics?

Metric	Definition	Tips
Share of Voice (SOV)/ visibility	How often you vs. competitors show up for queries your audience searches for	Check regularly – LLMs are constantly evolving, so SEO strategies that were successful 2 months ago might not work today.   Use AI Search Share of Voice as a relative (not absolute) metric – run queries you think your target audience uses through various LLMs to measure how often you and competitors appear. Use these results as a baseline that can identify opportunities for improvement and help track progress over time.
Sentiment	How customers feel about your brand – and how you are being perceived	Use AI to make sentiment analysis more quantitative – language processing tools can categorize brand mentions and track whether the sentiments behind them are positive, neutral, or negative.   Sentiment can indicate general brand perception – track whether AI tools present your brand the way that you want to be perceived. For example, a company that has recently pivoted to serve enterprise customers should track the extent to which it is being positioned as an enterprise, mid-market, or small business solution.
Traffic	Website visits from LLM citations	Remember that CTR is hard to track because LLMs site sources inconsistently – AI search tools are getting better about citing sources, but AI search CTR is more complicated to calculate because a click is often the last step in a multi-search process.

What tools can help you optimize your AI SEO strategy?

Current AI search tools only have monitoring capabilities – it’s not yet possible for brands to impact the results provided by LLMs. The closest you can come to “toggling” different factors is asking the tool itself how it’s choosing its answers and incorporating those insights into your AI SEO strategy, but LLMs aren’t necessarily able to provide accurate information about how they make decisions.

Specialized AI search optimization tools are emerging – highly regarded founders are leveraging expertise in adjacent spaces to create promising tools:

• EverTune – measures share of voice across LLMs and how brands are represented in those queries.   
• Poe – aggregation service that allows users to run the same query across multiple LLMs simultaneously. Companies can use Poe to get a sense of how they are being represented across the AI search ecosystem.  
• Air Ops – enables the creation of AI-friendly content at scale.  

Homegrown solutions can also help monitor AI search performance – use a platform like Zapier to measure ongoing performance by running queries into ChatGPT, or run 10 queries and feed them into a table in Clay that will explain your share of voice.

Learn about the tools and solutions that other people are experimenting with – many leaders in the AI and advertising spaces are sharing ideas and thoughts on LinkedIn. Take the time to read about what they’re trying and explore how their ideas might apply to your business.

Overall

How should your allocation of marketing spend change in response to AI search?

Reallocate some demand capture budget to PR and customer management – external voices now play a more measurable role in driving discovery. Reassess the impact of other top-of-funnel channels compared to PR and customer management, and test new budget allocations accordingly.

The largest shift should be in how you prioritize investing resources, not money – LLMs don’t let brands spend money to improve performance, so if you pull spend from Google display or Google search, there isn’t a clear channel to move it to. However, you should expect to spend more time developing relevant content and exploring answers to this problem in the future.

What are the most important things to get right?

Choose a clear starting point for how you want to incorporate AI – companies get overwhelmed by the vague need for “an AI strategy”. Start with a specific use case that will either help you create value for your customer or create efficiencies for your business. If you are concerned about why Google ads traffic has dipped and how you can improve your visibility across the tools customers are using to search for products in your category, it is worth investing time and energy in building an AI SEO strategy.

Remember that lower funnel activities still depend on traditional search – LLMs are most helpful in aiding discovery and helping customers identify the products that they want to learn more about. Once those customers become more intent on considering specific products, they use traditional search to learn more. Your SEO strategy must keep both types of search in mind.

What are common pitfalls?

Don’t make rash decisions about pulling spend from traditional channels – some companies are overreacting to the decrease in Google search volume by pulling spend. However, because there are no alternative places to utilize that spend, their full funnels are suffering. Begin to think about search more holistically but hold off on significantly shifting spend until you have tested and proven the viability of a new strategy.

Don’t be discouraged by AI search’s “awkward” growth phase – established companies and AI-first startups are all trying to figure out how AI search will mature. Patience, persistence, and a learning mentality are essential during periods of rapid change.