Ryan Walsh
Date: 02/09/2026

Many businesses worldwide want to be mentioned and cited on AI platforms such as ChatGPT and Perplexity.
However, many businesses cannot fathom why their business isn’t cited as a source on these AI platforms.

As a result, businesses worldwide now want to optimise their websites so they appear as citations.

This article will help you achieve that.

Table of contents:

• Which AI models do you want to optimise your website for
• We look at the common issues that render a business invisible within the AI answers and citations
• We look at how businesses can grab more visibility for their brand in the AI platforms
• We look at how businesses should audit their AI visibility

It’s no surprise that more shoppers are using, say, ChatGPT to find services and products.

This means, for example, a lot of companies will use ChatGPT to find out, for example, “who’s the best digital marketing agency” or “who’s best to optimise my e-commerce website”. A list will come back, but sometimes even established businesses won’t see their company cited.

AI systems need to find, understand your brand, then suggest you as the answer.

AI systems have a lot of choices to choose from. Businesses today therefore need to understand semantic SEO and entities.

For a long time, we have been talking about entities. Now, what an entity is in online marketing: it could be a business, an event, or, let’s say, a person. This idea emerged when Google started using what is called the Google Knowledge Graph.

So, it’s important to build a brand, even though that’s a rather vague term. However, by the end of the article, you’ll have a much better understanding of what we mean by entity building.

73% of B2B buyers now use AI tools

It has been cited by Averi research (March 2026) that when they analysed massive amount of AI citations, 680 million of them, that 73% of B2B buyers are now using AI tools. This means that ChatGPT and Perplexity is being used for purchase research. This again shows the power of AI in terms of how it’s being used to shape purchase makings decisions.

Higher conversion rates on AI

We have read a lot of articles around this, where it suggests that because of the conversational nature of AI platforms, the conversion rates are actually much higher.

For example, a comprehensive study by Loganix in 2026 the “B2B AI Buying Behaviour Analysis” that AI visitors convert more than Google’s organic traffic. For example, it was quoted that 14.2% of AI search traffic is the conversion rate, compared to 2.8%.

How to the AI models, such as ChatGPT form an answer

How Google’s algorithms work is they use ranking factors to place in order the organic results. However, when it comes to the AI systems, such as Perplexity, Claude, ChatGPT and Gemini, well this work completely differently. They instead synthesise the information, and they collect the answers from authoritative sources.

For example, if a website is a clear authority on a certain subject, let’s say family law or Swiss luxury watches, if that is an established and authoritative source, that will often be cited.

Similarities between AI models and SEO

It’s important to know that there are strong crossovers between GEO and SEO. For example, if a business has a really good content marketing, with strong E-E-A-T this will mean it rank higher organically and the written work is more likely to be cited by the LLM’s as well.

Clearly defined brand and entity recognition

This again goes back to the concept of Google Knowledge Graph. This is nothing new, in matter of fact, our SEO consultants were talking about this over 10 years ago. What basically should occur is that your business needs to define the brand, to define the entity.

For strong AI citation confidence, the business often needs to be established at offering helpful advice.

The brand needs to be seen as experts in their respective area. What this means is, the company needs to publish content marketing with high E-E-A-T score. The work needs to be helpful, the work needs to have backlinks, but also brand mentions on important websites.

The brand needs to be strong by completing the following:

• Authors of work need to be experts on offering that product or service
• Detailed answers through the content marketing
• High E-E-A-T score
• Semantic SEO, brand mentions + links

Structured data and Schema mark-up

JSON-LD schema markup helps inform the AI systems what the page is about. For example, using JSON-LD schema markup you can place machine readable text on the page. This allows you add for example article schema, product schema, and schema on the homepage for example to mark-up NAP information.

How the different AI models work is slightly different, however, the AI models in general like cross referenceable facts. They also like it when for example, the author is a genuine expert on that subject. Checking out whether an author is a genuine expert on the subject, that means checking for example LinkedIn accounts, plus building up knowledge on the author, by crawling and indexing perhaps a lot of the author’s previous work.

Its important to also consider page structure, good on-page SEO should be used. Then after each main heading, use a short and snappy answer, this is so that text can extracted and used as a direct answer to that question.

AI training data

The LLM’s such as ChatGPT are trained on training data, such as reading the pages on Wikipedia.

Now it’s been stated that 7.8% of all citations on ChatGPT originate from Wikipedia. This is why it’s super important to get your brand referenced on for example Clutch or Wikipedia, so that the LLM’s can obtain key information about your business.

Google’s E-E-A-T signals

We have written a tremendous amount on the subject of Google E-E-A-T. Basically, what Google’s algorithms want is for the work to clearly indicate that the author has a lot of experience, expertise, authoritativeness and trustworthiness.

Obviously, this doesn’t happen overnight, and this is why it’s important to have authors who can clearly show their experience. It’s also widely thought that the LLMs and AI like it when the information is backed by extensive market research and sources are cited. Also, when you have real customer reviews, once these accumulated over an extensive period of time this improves the trustworthiness aspect of the work.

Machine readable

For a very long period of time, Seo agencies would blog and write extensive advice on why a business shouldn’t have conflicting NAP signals.

Now this relates to local search engine optimisation. However, the same concept now applies to GEO or LLM optimisation. The reason being is you want your brand your business to be clearly understood by the LLMS. Now if for example, there’s another brand with similar name, this can cause confusion.

Again, businesses with messy NAP for example, this can be another point of confusion. For example, if the business name is typed differently on Google Business, to what it appears on LinkedIn, then its typed differently on the business directories, this can cause confusion not only to Google’s algorithms but also the LLMs.

Content should be written for answer extraction

What the LLMs want, including Google AI answers, is for an answer to be clearly extractable. Now what we mean by that is, you should structure the page, so that there are clear questions and answers. Now what this allows, is for the author to offer a very concise answer, which the LLM’s prefer. This means that the answer can then be extracted, and placed as for example a Google AI Overview answer. Now if the surrounding text, or the page is full of marketing waffle, this makes it harder for the LLMs to extract a very exact answer.

Therefore, it’s always best to write an exact answer and keep the answer on point.

Schema mark-up

It’s important to add schema mark-up to your website.

The two types of schema most commonly used would be organisation schema, which marks-up key information such as business name, URL, description of the services or products, founding data, location and which areas you serve.

Then there’s also schema for blog content as well.

Mentioned on trusted sources

This comes back to our previous point regarding Google Knowledge Graph and the semantic web. So, it’s important to improve the on-page SEO, however it’s widely thought to get the AI models to cite your work, you need to get your business cited on trusted websites.

For example, if your business has a lot of positive reviews on review platforms, this improves the trustworthiness of the business. If your business conducts market research and this gets mentioned on leading industry publications, this will be read by the LLMs. This means that the statistics or the data you have produced is more likely to be cited by the LLMs.

Crawlability

This is an important point, GPTbot, for example is used by OpenAI to crawl content. Now if you have written the robot.txt file to block that bot, then your business can’t then appear in the ChatGPT results. Therefore, you do have to ensure that the bots can crawl the work.

Kept updated

It’s widely thought that the businesses which keep their content marketing up to date, are more likely to see uplift in organic traffic and also referrals from the AI models, such as ChatGpt.

Does organic or traditional SEO work help improve AI visibility?

There are strong crossovers between organic or traditional SEO and GEO. For example, strong backlinks, plus brand mentions on leading publications, can help improve the E-E-A-T score for the business. That is good for organic and also for GEO as well. However, as this article illustrates, the content marketing must be high quality for it to be cited by the LLMs. Normally there’s a lot of time needed to demonstrate the authors are experts in their respective area.

How we can help

We are GEO, AEO and SEO experts.

We can help your company to generate more sales.

Call us today for more information.