Author: Digital Tailors
Date: 25/03/2026

LLM optimisation is the buzzword of 2026 in marketing circles.

You only have to glance at your LinkedIn feed if you work in marketing, and businesses are proclaiming they are the GEO, AI, and LLM optimisation experts.

LLM optimisation is therefore being advertised as the thing since sliced bread was invented!

Well, we will let you in on a secret: LLM optimisation does not mean reinventing the wheel.

It just means getting the normal SEO foundations right and bearing in mind a few more details as you go about creating content marketing for your business.

 

What is an LLM?

LLMs are simply large language models, and they basically read a businesses’ content marketing. This information, among with other signals, are used to then offer direct answers to questions.

If the LLM’s basically think a piece of content marketing is good, then they use it. This advice then gets sliced and diced with other information sometimes to offer a “synthesised” answer.

Therefore what is often produced is a synthesised answer, in which a business can have its information summarised and cited by a citation.

The citation sometimes offers a direct link to where the information has been read from.

Citations are the new battleground.

Okay, so SEO is not dead, far from it, so don’t think that SEO is no longer useful; it’s just LLM optimisation its also important. For those petrolheads out there, think of LLM optimisation like strapping a big turbo to your engine; get it right, and you can boost sales, just as a turbo boosts power.

 

Interpreting and understanding your brand

Since important updates such as Google Hummingbird, the Google Helpful Content Update’s, and Google EEAT- Google now better understands your content marketing.

A lot of top SEOs refer to this as the semantic web, basically understanding where your brand and products sit in terms of how important the business products /services are.

The semantic web, therefore, understands the wider picture, such as who the directors and who the key staff are, the awards the business might have won, how the product is different, and who the author of the content marketing is for example.

 

How can we better optimise our content marketing for the LLM’s?

Clear H1 and H6 hierarchy

It’s been the case, well for donkey years, that good page structure is important. This was important for organic SEO; it’s also important now for LLM optimisation.

Basically, whether it’s Googlebot or another bot that’s analysing the written work for LLMs, it just makes sense to better structure the content using H1-H6 tags.

Bullet lists

Try to make a point of using bullet points to convey what you mean quickly.

So, for example, if you’re talking about Google updates, that impacted this area of SEO in recent years, perhaps bullet-point them.

It’s thought that by doing this, Googlebot and the LLMs can quickly grasp what you’re talking about, if keywords are grouped in, say, bullet points. But obviously use white-hat optimisation methods should still apply.

 

Concise and to the point

Some copywriters, including us, do have a habit of waffling and going off topic sometimes.

This probably harks back to an era when SEOs globally were going down the road of writing longer and longer articles. Evergreen content marketing articles that were say in excess of 3,000 words, they are still useful, if the information is high quality.

However, with the LLMs, they do tend to like a very direct and precise answer to a question, this doesn’t mean articles must have a low word count, far from it. What it does mean however is that the page needs to be well structured, and set out. Think H1-H6, plus also use plenty of sub headings and titles, so the LLM’s can extract an answer.

Now, if the work needs it, it’s always a good idea to write a long article, just make sure the page structure is well thought out.

Therefore many of best digital marketing brains think that LLMs like to extract nice, to-the-point sentences.

This makes sense when you think about, say, ChatGPT or Google AI Overviews: normally, a short sentence or paragraph is paraphrased.

Therefore, if you make it hard work for the LLMs to extract a nice, to-the-point answer, well, they may go elsewhere to obtain that answer.
So, think about using short sentences, and to the point.

 

FAQ sections

This builds on our previous point about making it a piece of cake for LLMs to extract an answer.

If you answer FAQ questions in your written work, well, it just stands to reason that it makes it much simpler for the LLMs to extract a direct answer to a question.

Schema markup

Time after time, Schema is mentioned as being important to the LLM optimisation process.

So it stands to reason that this must be true.

Even if Schema doesn’t work in the context of LLM optimisation, it’s been useful for everything from local SEO through to e-commerce SEO, so it’s well worth using anyway.

Facts and figures

This again has its roots in traditional organic SEO, if you publish high-quality research, then you’re more likely to be cited by LLMs.

Many top marketing boffins have noted, for example, that Perplexity cites research from leading brands in its answers.

Therefore, why not quote leading research or conduct your own market research?

If you gain enough notoriety for a piece of in-depth research, then, from that, your work can be cited.

 

Google EEAT and authority in particular

Let us dispense a top piece of advice for you.

It’s simple, you know, sometimes the answer is right under your nose, well, this is true for LLM optimisation.

That’s the “Google EEAT guidance”.

In a nutshell, Google has told businesses to focus on providing experience, expertise, authoritativeness, and trustworthiness.

All super important.

However, if you were to ask us for our penny’s worth, we would say that businesses with the best and strongest backlink profile will do well not just in organic search, but also, you guessed it, in LLMs.

How you should not implement LLM optimisation

Using AI for AI

What do we mean by such a statement?

Well, some bright sparks might think that they are going to use AI to write the text.

Then, because AI has written it, surely AI will want to cite that?

We don’t think that will work; in fact, it’s a flawed marketing strategy from day one.

Google and LLMs such as ChatGPT, aim to cite the opinions, views, and advice from the best companies and voices out there.

Therefore, whether it’s a motorbike review or advice from a top solicitor, you have to become a voice worth listening to.

Now, that’s not going to happen if you’re using AI to write the content; it won’t be original.

 

Good technical SEO matters:

• Fast-loading pages
• You’re going to need a fast website, think super-fast.
• Don’t purchase low-cost website hosting that’s slow, plus make the web design super-fast.
• Good website design and architecture
• This is rather fancy marketing language for designing a really good, user-friendly website that basically offers all the advice the shopper want when buying a product or service.
• Don’t therefore have a ton of content on thin pages.
• Mobile-first performance

Most people today are time-poor, so this ties back to why you need a fast website.

If the websites are slow, then the bounce rate increases. However, mobile design is more important than ever. If you’re not managing your crawl budget, reducing how long the website takes to load, and offering a good UX, then all of these need to be worked on in terms of improving.

 

Schema and on-page SEO

Use all the normal on-page optimisation methods, such as H1-H6, good internal linking, and alt text. LLM optimisation might sound futuristic and super complicated, but it still has strong cross-overs with organic SEO, and this shouldn’t be forgotten.

 

What is LLM Optimisation, and how’s it different from SEO?

Our SEO agency explains how to improve your business’s LLM optimisation. We offer SEO, GEO and LLM optimisation advice.