AI search engines like ChatGPT, Gemini, Microsoft Copilot, and Perplexity.ai are rapidly transforming how people find information, evaluate solutions, and interact with brands. Instead of sifting through blue links, users are now having conversations with agents and making decisions before they even get to your website.
While Google still commands 89.71% of search market share according to BrightEdge’s Generative Parser (April 2025), monthly traffic growth is strong across AI platforms:
- ChatGPT: up 19%
- Perplexity: up 12%
- Claude: up 166%
- Grok: up 266% (early-stage spike)
The shift to answer-based discovery requires us to rethink how we measure visibility and optimize content performance, especially for enterprises.
When it comes to optimizing content performance for AI search engines, structured data is more critical than ever for these modern search features, according to Google during their Search Central Live presentation in Madrid. As AI becomes a core part of the search experience, the demand for precise, high-quality structured content is reaching new levels of importance.
Yet for enterprises, this shift introduces a new challenge: traditional analytics fall short when it comes to tracking visibility and performance across AI-powered platforms. To demonstrate value and stay competitive, organizations must rethink how they measure success, especially when AI referral traffic isn’t easy to access or attribute.
Why Do Enterprises Need to Pay Attention to AI Chatbot Visibility?
AI search isn’t just another channel, it’s an entirely new way your customers are discovering, evaluating, and interacting with your brand. Enterprises must prepare their content data to be accurately understood by these AI search engines, otherwise they risk losing visibility and falling behind more proactive competitors.
AI search engines act as filters of trust, selecting content that aligns with their understanding of entities, facts, and authority. Brands with clearly defined and semantically structured content are better understood by AI search engines resulting in visibility in AI-driven responses.
A recent BrightEdge study reported that over 53% of marketers are using two or more AI search platforms weekly, and that being cited in AI experiences is now a key performance signal. AI is no longer just a passive tool; it’s an active evaluator, interpreting intent, forming opinions, and determining which brands to engage with.
And good news! Structured data (aka Schema Markup) helps you show up in these experiences according to Google.
Adoption on these platforms is growing, however measuring the impact to your marketing funnel and how you are performing in these new experiences is not easy. Google does not currently provide AI Overviews or AI Mode metrics in Google Search Console, and there are no out of the box brand performance measures for ChatGPT. In absence of these resources, we’ve put together some approaches to measuring performance that are working for us.
How to Monitor Your Performance in AI Search Engines
Start with Manual Testing
Manual testing remains the most accessible way to understand how your brand is performing in AI search engine environments.
Start by prompting the chatbot with branded and non-branded queries that align with your core services, industries, and target audiences. For example:
- “What companies offer enterprise Schema Markup solutions?”
- “Who is Schema App?”
- “Best SEO platforms for healthcare marketers”
From here, ask follow-up questions like:
- “Where did you get that information?”
- “Can you provide a source?”
- “Is Schema App a trusted provider?”
This allows you to evaluate which links are surfaced for which queries and whether the information presented aligns with how you want your brand to be perceived.
To make your testing more strategic, consider:
- What are the top queries driving conversions to your site today?
- What are the key entities and topics your content focuses on?
- What types of questions align with your content and audience intent?
By performing informed testing based on these top queries, entities, and topics that you’ve identified, you can tailor your AI search queries and investigative work to cover what is most important to your organization. This helps evaluate not just if your content shows up, but how accurately it represents your brand and how effectively it supports your business goals. A best practice is to do this monthly so you can see how it’s changing over time.
What If the Information It Gives About My Organization Is Inaccurate?
If your testing reveals incorrect or misaligned information in AI search results, this is an opportunity to take action and look at the accuracy of your Schema Markup and its coverage across your site. Schema Markup ensures your content isn’t just written for human readers but also structured and understood by machines.
By implementing robust and dynamic structured data that clearly defines your organization, products, services, etc., you help AI systems ground their responses in facts you control.
Schema Markup empowers you to guide how AI interprets and presents your brand, improving both visibility and accuracy in these AI spaces.
Use GA4 to Track AI Referral Traffic and Performance
While, at this point, AI search engines don’t always pass clear referral data, there are several ways you can infer and measure traffic coming from these tools using Google Analytics 4 (GA4).
1. Monitor “Direct” and Referral Traffic Patterns
Some chatbot interactions, particularly those using embedded browsing like Bing or Perplexity, may show up under known referral sources (Perplexity.ai, Bing, Gemini). However, many visits appear as “direct” traffic due to stripped referrer headers.
Look for:
Spikes in direct traffic to specific pages shortly after testing prompts.
Referrals from new or unexpected domains.
Changes in source/medium behavior in your GA4 acquisition reports.
Tip: Use the GA4 report: Reports > Acquisition > Traffic acquisition > Session source/medium and scan for new AI-associated domains. See example below:
Every session source/medium type being tracked in GA4 will come up in the report shown above. From there if you wish to isolate individual AI sources, you would either create a secondary filter isolating all the sources you are interested in (we recommend using a custom regex), or you can search for each AI search engine source individually using the search bar at the top.
As of early 2025, we’re seeing that traffic from AI search and chatbots tends to be higher quality with longer average session durations compared to the site average. You can read about what we’re seeing on Schema App in our article titled, Navigating AI Chatbots and AI Overviews: Key Observations and Strategies.
2. Performance by Query Intent
To better understand how AI search is impacting your website, it’s essential to evaluate performance through the lens of query intent. This means moving beyond aggregate metrics to examine why users are arriving on your site and what they’re looking for.
Start by asking yourself:
- Are your brand, product, or service pages ranking for high-conversion (transactional) queries, or are they showing up for more informational searches?
- Which types of questions are leading users to your site? Are they researching solutions, comparing vendors, or ready to take action?
- You can figure this out by assessing what queries your AI-ranking pages are achieving in Google Search Console or platforms like SEMrush, and also by inferring based on the types of pages people are landing on within your referral traffic sessions found in GA4 (see example below).
Use a mix of test prompts across both intent types:
- Informational: “What is schema markup?”, “Best practices for healthcare SEO”
- Transactional: “Best schema markup solution for enterprise”, “Request a demo Schema App”
Analyze which intents are driving traffic and engagement:
- Do your AI-driven mentions skew toward awareness or conversion? Ie. If most of your traffic goes to blog pages, it’s likely awareness/informational intent, if it’s converting pages like pricing or solutions, it’s likely conversion. As you can see in our example above, our AI referral traffic is mostly in the awareness and informational intent category.
- Which intent group leads to higher quality sessions (longer duration, deeper engagement, higher conversion rates)?
From these insights, you can:
- Refine content to better match the dominant intent group.
- Identify gaps where you may need more transactional or informational content.
- Improve internal navigation and CTAs to support user journeys that begin in AI platforms.
This query-level analysis ensures you’re not just monitoring visibility, but aligning your content strategy with the reasons users are arriving, helping you prioritize the traffic that matters most.
3. Evaluate Traffic Quality
Success metrics are shifting. AI is driving more impressions but fewer clicks, meaning it’s essential to move beyond just sessions and focus on quality indicators, such as:
- Average engagement time per session
- Pages per session
- Scroll depth (if configured via GTM or GA4 events)
- Conversions (e.g., form fills, demo bookings, downloads)
Tip: If you notice low-quality traffic from AI sources, it may suggest misleading citations, brand confusion, or misaligned content.
The good news is that if your content seems to be misunderstood by AI search engines, there’s a solution for that!
How Schema Markup and Content Semantics Drive AI Performance
If you want to improve your visibility in AI search, optimizing your website for machines, not just humans, is essential. AI systems rely heavily on understanding context, relationships, and factual accuracy. This is where semantic SEO and structured data play a critical role.
1. Implement Robust Schema Markup
Schema Markup plays a foundational role in transforming your website content into data that machines can interpret with clarity. To ensure AI systems can fully understand and evaluate your site, use structured data that reflects the meaning and relationships behind your content.
Google and Microsoft confirmed in March 2025 that their LLMs are using Schema Markup to ground AI-generated answers. This tells us how your content is contextualized now directly impacts whether it’s visible in AI search.
Organizations should prioritize marking up the content that supports a customer’s journey, from discovery through to conversion. This includes:
- Informational content that answers key audience questions.
- Mid-funnel content that establishes trust and credibility.
- Bottom-of-funnel assets like product, service, or conversion pages that drive action.
When executed properly, Schema Markup can signal authority and relevance to LLMs, enabling them to more confidently include your brand in AI-generated results.
But Schema Markup on its own isn’t the end goal. Its value multiplies when your markup is connected through a semantic network of entities, forming relationships that clarify not just what your content says, but what it means within a broader knowledge ecosystem.
2. Internal and External Entity Linking to Clarify Content
After implementing Schema Markup, you can take it a step further by internally linking the entities you identified within your Schema Markup to other known entities across your site, as well as externally linking to authoritative knowledge bases across the web such as Wikipedia, Wikidata, Google’s Knowledge Graph, and even official social media profiles.
This is known as Entity Linking, and doing it both internally and externally brings depth and disambiguation to your structured data.
Entity Linking helps AI systems connect the dots between your content entities and known terms, enhancing their ability to infer meaning. By anchoring your entities to trusted external sources, you provide context machines can rely on, improving your chances of accurate inclusion in AI-generated answers.
Tip: Use “connector” properties such as “sameAs” in your Schema Markup to clarify these relationships. This ensures the AI knows whether you’re referring to Apple the fruit or tech brand, Mercury the planet or the element, etc. These distinctions are critical in an AI-driven landscape where precision equals authority. Learn more about how to build semantic Schema Markup here.
3. Brand Mentions and Coverage Reporting
We’re hearing from many Schema App customers that brand mentions in AI search are becoming a high priority. In this context, a brand mention refers to:
- Your brand being named in chatbot responses (even if not linked, making it hard to currently track).
- AI referring to your company as a provider, solution, or authority.
- Mentions that carry sentiment, endorsements, or descriptions of capabilities.
This is now a visibility signal in the eyes of LLMs. Just as backlinks influenced traditional SEO, mentions and citations in AI help determine brand authority.
Coverage also matters. It’s not just about showing up – it’s about being accurately represented. Enterprises need structured data that:
- Dynamically matches real-time page content.
- Is complete and accurate across entity types.
- Is updated at scale and validated regularly.
This ensures AI systems cite and describe you the way you want to be seen.
Learn how to navigate AI search as a Digital Marketer
4. Build a Content Knowledge Graph
As your Schema Markup becomes interconnected through both internal and external Entity Linking, you’re laying the foundation for a Content Knowledge Graph – a reusable, structured data layer that reflects how your website content is organized conceptually and semantically.
A Content Knowledge Graph maps your topics, entities, and relationships using standardized vocabularies like Schema.org, and when expressed as RDF triples, it allows for precise querying, data reuse, and advanced AI interpretation.
At Schema App, we build Content Knowledge Graphs for our customers through Entity Linking using our Highlighter tool, which applies semantic structure to content at scale.
Ultimately, building a Content Knowledge Graph isn’t just about SEO; it supports intelligent AI interactions, drives data consistency, and ensures your content is ready to perform in emerging search experiences.
Turning AI Visibility Into Business Impact
AI search engines are no longer fringe tools – they’re becoming the front door to digital discovery. For enterprise brands, understanding how your content is surfaced (or overlooked) in AI chat experiences is now critical for growth and competitiveness.
By combining manual testing, GA4-based tracking, and semantic optimization – including Schema Markup, Entity Linking, and a Content Knowledge Graph – you can measure, enhance, and ultimately own your brand presence in the age of conversational search.
Want to ensure your website is AI-optimized?
Get in touch with us to learn how we help enterprises implement structured data at scale and align their content with the future of search.
Martha van Berkel is the co-founder and CEO of Schema App, an end-to-end Semantic Schema Markup solution provider based in Ontario, Canada. She focuses on helping SEO teams globally understand the value of Schema Markup and how they can leverage Schema Markup to grow search performance and develop a reusable content knowledge graph that drives innovation. Before starting Schema App, Martha was a Senior Manager responsible for online support tools at Cisco. She is a Mom of two energetic kids, loves to row, and drinks bulletproof coffee.