How visible is your brand in AI search? If you’re not showing up in responses to relevant queries, the problem probably isn’t poor content (although it doesn’t help). Instead, the various AI systems might not have a clear enough handle on your brand. That’s not a ranking problem. It’s an entity problem – and it fundamentally changes what “being found” online means.

Yes, search marketing has always been about being found. But as the tools we use change and behaviours evolve, how you get found has shifted.

Since the beginnings of SEO, visibility in search has hinged primarily on keywords, returning whichever content contains the same strings of text. But AI systems like ChatGPT, Perplexity, Gemini, Claude and even Google’s AI Overviews don’t work like that. Never mind words; LLMs are far more concerned with understanding the entity each word represents.

Simply put, an entity is a distinct, named thing that AI and search systems can identify, categorise, and connect to other things. Organisations, people, products, places, concepts – these are all entities. Your brand is one too.

As such, entity SEO is essential to ensuring LLMs recognise, understand and trust your brand – particularly in relation to relevant topics. Unless LLMs can confidently identify what your brand is, what it does, and how it relates to other relevant entities, they won’t surface you in answers, summaries, or recommendations.

But while entities have taken on a whole new level of importance with the rise of generative AI and LLMs, entity SEO is not exactly new.

From keywords to entities

The difference between keyword SEO and entity SEO is the difference between matching and knowing.

We’re all familiar with keyword-based systems, returning whichever content contains the same strings of text. How they rank these pages may have become far more sophisticated, but the word is still – ahem – the key.

However, keywords often lack the necessary context. Do you mean apple the fruit or Apple the billion-dollar tech company? When you type in blade runner, do you mean the classic 1982 sci-fi movie or disgraced Paralympic sprinter Oscar Pistorius?

Many things might share the same keyword, while representing very different entities. But even if the search engine or LLM understands that you mean the Steve Jobs company, what does it really “know” about Apple beyond those five letters?

A lot, actually.

The foundations for entity SEO were laid back in 2011 when Google, Bing, and Yahoo collaborated to standardise structured data mark up with the launch of Schema.org. Webmasters could mark up their content with additional information to explicitly tell web crawlers what their content was about.

In 2012, Google introduced the Knowledge Graph. This was Google’s first step away from pure keyword matching and towards, as Google engineer Amit Singhal famously put it, understanding “things, not strings”. A year later, Google’s Hummingbird update improved the search engine’s ability to understand the intent behind natural language queries by drawing directly on Knowledge Graph relationships.

Google went on to launch RankBrain in 2015, using machine learning to further deepen the engine’s ability to interpret context and relationships.

However, entity signals were still just one factor among many. Developing a solid knowledge graph for your brand could certainly help improve rankings, but it wasn’t necessarily a deal breaker.

Generative AI and LLMs have changed all that. When constructing coherent, accurate, and relevant answers, these systems are far less interested in keywords and far more dependent on the relationships between entities.

In AI-driven search, your visibility depends on recognition and trust, not just rankings. And if AI can’t confidently identify your brand, they simply won’t surface you.

This raises the stakes for brand visibility, particularly as AI-driven discovery increasingly bypasses traditional click-based journeys altogether. In traditional search, visibility is about degrees, rising or falling with your rankings. You might rank low down on the first results page, or even somewhere on the second or third page, and still capture some clicks from those users willing to scroll further to find what they want.

But in AI, brand visibility is far more black and white. You’re either cited in an AI-generated response or you don’t exist. If the LLM determines your entity is less relevant to the query than two or three others it decides to cite instead, no amount of scrolling or clicking will cause your brand to suddenly appear.

How AI systems understand your brand

In many ways, entity SEO is no different to brand building.

Traditional marketers will likely be familiar with The Marketing Rule of 7, a general principle which suggests that a potential customer is unlikely to make a purchase until they have encountered a brand at least seven times.

If they come across your message only once or twice, they’re unlikely to bite. But if they hear about you from seven or more authoritative sources – a colleague recommends you, a trade publication mentions you, a Wikipedia entry details your brand’s history – most people would consider your brand established enough to trust what you have to say.

LLMs basically do the same thing. They triangulate across a wide range of signal sources – Google’s Knowledge Graph, Wikidata, Companies House records, industry directories, news mentions, and more – to build a picture of your brand and ascertain how authoritative it is in its domain. The more trust signals there are, the stronger and the more credible your entity becomes, increasing the likelihood of LLMs mentioning or citing your brand in response to relevant queries.

Where most brands fall down

Many brands aren’t fully aware of how their entity appears to LLMs – not because they’ve done anything wrong, but because entity SEO wasn’t necessarily prioritised until more recently.

When SALT audits a new client’s entity signals, the most common issues we see include:

Inconsistent brand naming

If your company is referred to differently across your website, social profiles, industry directories, and news mentions, you’re sending conflicting signals. Machines are very literal, so even small inconsistencies in punctuation and capitalisation can weaken an LLM’s confidence in the identity of your entity. This can also happen if your trading name differs from your registered company name and you haven’t made the relationship between these two entities clear wherever possible.

Missing or weak organisational schema

Structured data clearly and unambiguously tells LLMs and search engines who you are: the brand’s name, address, parent company, key services, and so on. Some LLMs absorb this information directly; others don’t. But having this information correctly structured and available across the web still matters.

A fragmented digital presence or contradictory messaging

If your content tells one story but your wider web presence tells another, LLMs will have a harder time building a confident picture of your brand. Everything your brand publishes should reinforce the same core topics and expertise – particularly where they relate to products or services you offer.

Few authoritative citations

If the only mentions of your brand come from online sources with similarly weak or less trusted entities, LLMs are unlikely to take much notice. The volume of backlinks and mentions, reviews and recommendations, is far less important than the quality and authority of those sources.

How to strengthen your entity signals

Don’t panic. Entity SEO isn’t a separate discipline requiring a whole new strategy and investment. It taps into many of the same brand building practices you’re probably already doing, but with an increased focus on how AI systems interpret these signals. All the usual levers come into play, including clear messaging, content strategy, technical SEO, and digital PR.

Some specific areas to focus on include:

  • Standardise brand naming across every touchpoint – website, directories, social profiles, PR activity – so that LLMs consistently encounter the same brand identity.
  • Strengthen organisational schema so machines can clearly identify who you are and how you connect to other known entities.
  • Invest in digital PR to generate authoritative citations. This might include press coverage, industry mentions, forum discussions, your Wikipedia presence.
  • Build a structured content architecture that clearly maps your brand to the topics and entities you want to be known for. Try to avoid straying into unrelated territory that might send contradictory signals.

While there are a few tools out there you might use to view your brand’s knowledge graph to get a sense of how your brand entity is currently perceived. However, a quick and dirty diagnostic test might be to simply ask each AI platform what it understands about your brand entity. How it responds should hopefully give you a reasonable idea about how well the LLM currently understands your brand.

Of course, this won’t tell you how your brand entity stacks up against your competitors.

A proper audit will usually take a bit more time and expertise. For one thing, it’s worth starting out with a workshop to clarify, agree, and document what your ideal knowledge graph should contain – which topics and relationships your entity should be known for – and comparing this with your competitors to identify any gaps or weaknesses to address.

The brands AI knows will be the brands customers find

While entities have been a part of search for years, they weren’t the primary driver of brand visibility. But with the arrival of AI-driven search, entity SEO has become a key part of any brand visibility strategy.

However, this isn’t just another technical project you can hand off to the relevant team and forget. Entity SEO spans the code and the content, owned and earned, marketing and PR, onsite and offsite … in fact, just about anything that relates to your brand online.

These things already mattered, of course. What entity SEO does is approach every element of your online brand with greater strategic intent, ensuring they all tell the same compelling story.

SALT helps brands build entity strength and improve visibility in AI-driven search. If you’d like to understand how your brand currently appears to AI systems, get in touch: [email protected]