Most CMOs looking at their analytics today will come to the same conclusion: AI isn’t driving enough traffic to matter. At a surface level, that seems rational. Why shift budget or strategy toward a channel that barely shows up in your reporting?

The problem is, you’re measuring the wrong thing. AI-driven search is already shaping buying decisions, long before a click ever happens. Prospects are researching categories, comparing options and forming preferences inside AI-generated answers – often without visiting a single website that might show up within your standard analysis platforms.

That means the visible traffic in your analytics is only a fraction of the real impact.So if you’re relying on analytics alone, it’s easy to draw the wrong conclusion.Think of it like an iceberg. Above the waterline are the AI referrals you can actually track. Beneath the surface are the countless AI-generated conversations that never result in a click because the user already got the information they needed – or because your brand was never mentioned in the first place.

And that’s the real issue. Most brands currently have little visibility into how AI platforms represent their products, expertise, or competitors. There’s no reliable way to measure many of these interactions today, but that doesn’t make them any less influential on pipeline, perception, or purchasing decisions.

This isn’t a future trend. It’s already changing how people research and buy.

If your strategy is based only on the limited AI referral data you can see, you risk underestimating a shift that is already reshaping online visibility.

Don’t make the mistake of basing your marketing and SEO decisions solely on the extremely limited AI data you can see. Continue full steam ahead without changing course, and that vast, hidden bulk can still rip a hole in your visibility strategy and sink your business.

Just ask the Titanic.

Rankings are no longer the whole story

For more than two decades, the logic of search marketing was straightforward. Rank higher, get more clicks, generate more leads. CMOs understood it. Boards understood it. It worked.

AI has disrupted that comfortable model. This isn’t just a shift in how search works, it changes how demand is captured. If clicks decline but decisions are still being made, then visibility needs to be measured in influence, not traffic.

I’m not saying rankings no longer matter. Things haven’t changed that much; at least not yet. But traffic from organic search is definitely changing.

These days, even the number one ranking in organic search isn’t the prominent, highly visible position it once was. Do a Google search today, and the top result might be buried below a lengthy AI Overview and a bunch of paid ads. Why should the user scroll to find and click on your link when the AI Overview has already answered their question?

If you really want your brand to be right at the top of the SERPs, you need to be cited in that AI Overview.

However, it would be wrong to think of AI citations purely as a new way to rank in search. This isn’t about being number one. The importance of citations goes far beyond your brand’s position on the page in a single channel.

A citation is what happens when an AI platform (ChatGPT, Perplexity, Claude, Gemini, etc.) or an AI-powered search feature (Google’s AI Overviews or AI Mode, for example) generates an answer that names your brand as a relevant, trusted source. It may include a link back to your site, or it may simply be a mention conveying valuable messaging.

If your brand isn’t cited in AI answers to relevant consumer queries, you are excluded from consideration at the very point in the journey where decisions are being shaped. You’re effectively invisible. If you’re not cited in AI answers, you’re not part of the conversation. End of story.

Instead, citations are about building trust and influencing decisions. When a buyer types a brand- or product-related query into ChatGPT, for example, they don’t necessarily jump straight to purchase. Chances are, they’re using AI to help them on a journey of discovery, research, and consideration. They might explore your category, shortlist their options, and weigh up alternatives without a single click registering in your analytics. Once they’ve made a purchasing decision, they might click through to the relevant brand to complete the transaction. If that brand is yours, your dashboard might finally register a single click. I say “might” because OpenAI is already integrating with e-commerce brands like Etsy and Shopify to offer users instant checkout options, allowing customers to hit a buy button on an AI recommendation and complete a transaction without visiting the retailer’s website even once.

Measuring what you can’t yet see

There are two reasons why so many AI conversations are hidden from your analytics:

  1. Tools like Google Search Console haven’t kept up with the rapid changes in both technology and search behaviours
  2. Data from the various AI platforms isn’t currently shared externally with these tools like Google Search Console.

Native citation reporting from the major AI platforms is coming – Google has already begun laying the groundwork – but it isn’t here yet in any meaningful form.

This lack of measurable data creates a new challenge for CMOs. You’re still expected to measure performance, justify investment, and explain pipeline trends, but a growing part of the buying journey now sits outside traditional attribution.

In the meantime, there are three approaches you can use to at least get a sense of whether LLMs are taking an interest in your brand and your content.

  • Monitor AI crawler and user-agent activity on your site: With SALT’s Cloudflare methodology, we can visualise how often AI systems visit your website to assess your content. While each visit doesn’t equate to a citation, consistent crawling is a strong indicator that your brand is at least being considered as a potential source.
  • Use AI visibility tools: There are some AI analytics tools out there, but they’re still largely keyword-dependent and imperfect. While they can’t track actual conversations, they can simulate queries according to your chosen topics to test how often each platform cites your brand in comparison with your competitors.
  • Examine AI-referred traffic conversion rates: Visitors who click through to your website from AI citations are more likely to be deeply engaged with the topic or further along the decision-making journey. If traffic from AI converts at a higher rate than typical organic visitors, it would suggest your AI citations are working.

While these methods can’t provide you with complete visibility, doing nothing is sure to leave you blind. So, pick a tool, establish your baseline, and track your direction of travel. You’re looking for trends rather than absolute numbers – and the earlier you start, the more meaningful that trend becomes.

How to earn more citations

If citations shape decisions before a click, then this isn’t just about content optimisation. It’s about ensuring your brand is consistently included in the answers buyer’s trust.

A content strategy built on listicles and keywords won’t cut it. If you want LLMs to determine your information is valuable enough to cite in response to the most relevant queries, you need original, highly relevant content – packed with information, insights, and genuine expertise the LLMs will struggle to find anywhere else.

In practice, this comes down to three priorities:

Technical foundations

If AI crawlers can’t access, parse, and trust your site, nothing else you do will matter.

Page speed is a good example. Our own research found that slower-loading sites were less-likely to be cited in Google’s AI Mode. Of course, all the usual basics apply; clean crawlable URLs, logical site architecture, no dead ends or crawl traps, and schema markup that helps AI systems understand your content’s context and structure.

Content clarity

LLMs don’t read content like you or me. Instead of ingesting entire webpages, LLMs break content into smaller chunks or blocks of text, before identifying and extracting only those chunks containing the information it needs.

This is called RAG chunking, often just referred to as “chunking”, the details of which are far too technical to get into here. But what it boils down to is this: If you want LLMs to identify, extract, and accurately paraphrase information taken from your content, you need to write and present key facts and figures in such a way that each chunk could potentially function as a standalone, self-contained answer. In practice, this means short, focused paragraphs, descriptive headings, and schema markup to provide extra context.

Entity authority

An LLM doesn’t assess each individual page or content asset on its own merits. It also considers what it knows about your brand, and whether you’re a recognised authority on a given topic. The LLM builds this picture of your brand entity from multiple signals, including the services you offer, topics you cover, mentions, links and recommendations, as well as your connections and relationships with other entities.

I’ve covered the mechanics of entity SEO in a previous article. However, the short version is that you can use strategic content, clear messaging, and digital PR to create a consistent brand presence that LLMs understand and trust.

The iceberg is out there

Citations are rapidly becoming the new mechanism through which visibility translates into influence. Buyers are asking nuanced, conversational questions at every stage of the buying journey. The brands that get cited are the brands buyers trust.

This changes what your growth strategy needs to achieve. It’s no longer enough to target the top of the funnel with broad awareness content. Your brand needs to be present and credible across the full decision-making process, or risk being excluded from consideration entirely.

The advantage is already clear. Early movers are shaping decisions now, reducing their reliance on paid channels later, and building a more resilient pipeline as a result. Brands that wait will find themselves competing for demand that has already been shaped elsewhere.

The iceberg might not be visible, and it certainly won’t announce itself. But it’s out there, influencing outcomes beneath the surface. The time to change course is now.

Not sure whether your brand is being cited or excluded? We’ll show you where you stand, what’s driving it, and what to do next. Get in touch with the SALT team to discuss your organic search visibility.