What is GEO and How to do it The Right Way

The GEO Definition

In part one of this series we explored The Rise of the Machines and why LLMs have become our Trusted Buying Companions. The article concluded that trust allied to convenience is the underlying power axis behind the meteoric rise of LLM adoption in buying journeys and it is the more fragile and volatile trust component that will ultimately determine their long-term future.

In this second part of this three part series we introduce the concept of GEO, what strategies and tactics we see emerging in a highly nascent field and endeavour to provide a solid best practice framework for implementation.

GEO – or Generative Engine Optimisation – is the online practice of improving a brand's visibility in LLM (Chat GPT, Gemini, Perplexity etc.) responses to prompts. It has been referred to in other terms. AEO – Answer Engine Optimisation – is used commonly although typically more in the context of Google AI Mode's shorter form responses. LLMO (Large Language Model Optimisation) and GSO (Generative Search Optimisation) have also been used however for the purposes of this article we will adopt the term GEO which is fast emerging as the industry standard definition.

GEO Is Just Like SEO Right?

Well the technical answer to this question is "sort of…but not really". It is true that there are SEO and GEO practices that overlap. Structural and technical changes to the website, such as schema mark ups (a code added to the website to help search engines, and AI systems, understand content better – reference this article for more information: What is Schema Markup?), page load speed and E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) signals have comparable impacts when it comes to LLMs citing a website versus traditional search engines locating a website.

At this juncture it is worth being super clear on the definition of citations versus (brand) mentions or references since both are key metrics when it comes to GEO performance tracking. Citations are source links – they are where the LLMs go to find content in response to user prompts. A brand mention (or reference) refers to whether your brand name was used in the response the LLM gave to the prompt.

This distinction, and then how businesses apply it to their GEO performance tracking, is extremely important to highlight. Citations are often reported by agencies or digital teams as the lead indicator however given LLMs tend to defer more to third party domains over first party (your own website) and up to 93% of LLM experiences end without a click (Exposure Ninja, 2026) this is typically misguided (unless the brand in question happens to be a media company for example, where citation volume infers brand relevance).

Brand Mentions are more often than not the lead indicator – not just volume which is typically captured as a % brand visibility score in GEO data platforms (number of brand mentions divided by number of prompts run through the LLMs) but also Brand Sentiment. In other words how positively or negatively your brand is referenced in the response content. Brand Position is important (how high up your brand features in an LLM response) but needs to be delivered in conjunction with "fit" i.e. your brand showing up in the right context in response to prompts.

Brand Accuracy is also emerging as a metric (how true is the response in relation to the product or service actual capabilities) however both sentiment and accuracy measurements require a certain level of language nuance from the algorithms which can lead to misrepresentation. These metrics should be treated with caution and interrogated deeply at this stage.

Back to SEO vs GEO and let's stay on the topic of metrics. The SEO primary outcome and objective is clear cut – traffic to website. This is delivered via the binary metric of ensuring your brand is featuring top of rankings in Google search. And with SEO ownership typically sitting with an individual whose responsibility is 100% SEO, the discipline is largely contained.

Let's compare this with GEO. We have already established there are multiple GEO metrics. In addition there isn't a direct and linear attributable outcome such as clicks to website as we've also established. We would expect inbound demand, site conversion rates, brand KPIs (key performance indicators) and overall revenue to all improve via successful GEO however these are hard to directly attribute. Then there is the ownership and delivery of GEO which is highly fragmented across organisations in comparison to SEO. At lomo ai we call this operational challenge "the fragmentation problem". Let's elaborate on this:

For GEO to be successful, the Brand team has to ensure fundamentals such as the value proposition are precise and up to date to make sure the brand is consistently referenced in the right terms by LLMs in response to the prompts that are relevant. It is worth noting that the LLMs care less about the artistic side of brands (look, feel, tone of voice etc). They care about the hard science – is your product or service differentiated vs the competition, how relevant is it to your audience and why should they believe you.

Successful GEO requires inspirational thought leadership to be generated and translated into various forms of content that is then published in the right media to be cited by LLMs. Paid media and earned media (typically PR) strategies need to be aligned and supported by GEO analytics to build plans, place content and measure effectiveness. The brand website (Owned Media) needs to be enhanced for GEO, with some SEO crossover, but other unique requirements – see later in article. Social channels (such as YouTube, Instagram, LinkedIn) are increasingly being cited by LLMs whilst Community Platforms such as Reddit and Quora are often deemed credible sources of content by LLMs due to the perceived neutrality of discussion forums. Finally GTM (Go to Market) frameworks need to be adapted to effectively integrate a GEO strategy as effective deployment drives an inbound intent motion whilst a number of sectors (notably in B2B) are heavily geared towards outbound (we will explore this further in Part 3 of this article series: "The AI era of Customer Buying and the Future of Growth".)

It is perhaps this high degree of operational complexity that demonstrates the biggest divide between SEO and GEO. SEO is largely binary: SEO Lead + Site Analytics + Site Technical and Structural (including Content) changes = effective SEO. GEO on the other hand, certainly with how the majority of organisations are structured today, is inherently disjointed with (in an extreme example) up to five agencies and five in house teams that need aligning behind a single common goal.

Primarily for this reason, and the significant use of LLMs now in buying journeys (reference Part 1 of this article series: The Rise of the Machines: Why LLMs have become our Trusted Buying Companions), GEO cannot be considered a search or visibility tactic. Moreover it must be implemented as a growth strategy by organisations and sat with growth leaders (Chief Growth Officers, Chief Revenue Officers, Chief Marketing Officers) to operationalise and coordinate effectively with the right level of accountability. We will also explore this further in Part 3.

There is a deeper, more underlying point of variance between SEO and GEO which we can trace back to Part 1 of this series and the narrative surrounding trust. For many of us, LLMs have become trusted research and buying companions. They have become heavily intertwined in our personal and professional lives and our increasing reliance on them has helped forge a trust bond. The LLM experience is immersive as they guide us through sometimes complex and high risk investment decisions. Compare that to the traditional search experience which is highly functional – indeed in Part 1 we highlight that search engines are often now used as a lazy URL entry bypassing tool. Research by SparkToro reveals that 33% of web searches are made simply to find a specific website. The experience is concise and pragmatic. The user engagement doesn't happen at this point in the journey, it happens at the destination website.

What does this mean? Effectively it means when we employ GEO there is more on the line. There is more on the line for our brand and our business. Because if we, or the LLM, gets it wrong there is a de facto erosion of trust by association to an experience we have forged a bond with and are emotionally invested in. And remember once content is out there in 3rd party publications, it's out there and very hard to retrieve or amend – that is why the brand and content axis is integral to GEO. Users accept with traditional search it is a means to an end, that for the most part sponsored links are a promotional tool that should or shouldn't be avoided and that top of rankings doesn't necessarily mean the website is the best fit for their needs. Trial and error is accepted, and expected, and brand judgement takes place on the website not in Google.

This leads us nicely on to doing GEO the right way. But before getting on to this, maybe we should start with doing GEO the wrong way.

Doing GEO the Wrong Way

Learning from SEO

Continuing the SEO theme, there is a useful frame of reference for this appraisal. Black hat SEO was a phrase coined in the early days of traditional search for tactics that attempted to game or manipulate the Google algorithms to favour one brand over another.

In the late 90s and early 00s the algorithms were simple and relied heavily on direct text matching and link counts. This allowed digital marketers to easily game the system however by the mid-00s most of the tactics were explicitly banned by Google.

Here are some examples of black hat SEO tactics and how Google countered:

The penalties for such black hat SEO activities are now severe including:

Algorithmic Devaluation: The site instantly loses 50% to 95% of its search traffic as the automated algorithm stops counting the manipulated signals.

Manual Action Bans: Human reviewers at Google issue a formal penalty. This can lead to complete de-indexing, meaning the website is entirely removed from search results.

Fast Forward to GEO

We are seeing similar black hat patterns emerge with GEO. Instead of keyword stuffing there is AI phrase stuffing using semantic triggers and repeating exact technical phrasing that forces an LLM's attention – to the point where the copy itself makes no sense and loses credibility. For Cloaking we are seeing LLMs served a highly structured, clean, citation-dense research page whilst a human visiting the exact same URL is shown a standard, low-value affiliate landing page or an entirely different layout. And instead of Link Farming, some agencies are mass-publishing thousands of synthetic articles across the web, creating fake personas and surveys, in order to create an artificial consensus that LLMs identify as authoritative. In addition, community platforms such as Reddit have seen a rise in fake personas infiltrating discussion forums in an attempt to influence the narrative – an extremely high risk tactic that will lead to platform bans as well as LLM penalties.

Will these tactics win out? In short "no". The platform algorithms now are infinitely more powerful than they were 25 years ago. The war has already been waged by the tech giants on such activities because they undermine their own credibility so the battle lines are already set as we shift from the Digital era to the AI era of search and research. And as we've already established, the trust association with LLMs versus traditional search runs so much deeper, that a breach of this trust will be more damaging for the LLM (and the brand) in question.

Beyond the risk of algorithmic penalties and fines, there is a further watch out for brands that attempt black hat GEO which is even if they succeed short term in increasing visibility in LLM responses, if their brand is misrepresented in any way (which is more than likely since "black hatting" is inherently scattergun), those increases in mentions will largely be nullified when the entire customer journey is taken into account. Once again we are back to trust. A common black hat activity is to AI phrase stuff with every possible benefit, feature and capability in the hope that if this is highlighted in a prompt, the LLM will surface the brand as a potential option. However even if this works, the trust contract between the brand and the customer will break at the point the latter visits the website (which is still highly likely to happen in the journey even if we established it's less likely to be from an LLM direct click) because there is every chance there will be nothing or very little on the site in relation to the stated benefit. And as we have already established, because the user is heavily invested in the LLM experience and their trusted guidance, the retribution on the brand in question will be so much greater than false advertising or sponsored ranking elevation with a traditional search link.

Put simply, black hat GEO just isn't worth it. There are some grey areas (termed unsurprisingly grey hat GEO!) which push the boundaries such as synthetic corroboration (manufacturing the appearance of independent third-party consensus across owned or paid properties) and citation laundering (creating the illusion of multiple aligned sources that all trace back to a single origin). Once again the payback on these activities is highly questionable. There is enough in white hat tactics and in the 6 interconnected strategic pillars of GEO (lomo ai | 6 interconnecting strategic pillars of GEO) to sustain a highly effective, lasting and ethical GEO delivery plan.

Doing GEO the Right Way

GEO Tactics

Through this article we have touched on both the white hat tactics (the opposite of black hat and therefore fully legal and ethical) and strategic pillars that sit under the umbrella of "doing GEO the right way". Let's start with white hat activities. As previously mentioned, there are overlaps here with white hat SEO such as schema markup, page speed and E-E-A-T signals (strong FAQ and About Us pages help here) but there are also unique white hat GEO tactics such as Answer-format content structuring (Writing content specifically to match how LLMs construct responses – this is an excellent article that goes into more detail: Generative Engine Optimization (GEO): A Practical Guide | Reply) and Entity reinforcement across the open web (deliberately seeding consistent facts, figures, and brand descriptors across multiple independent domains so LLMs encounter the same information repeatedly).

GEO Strategy

The strategic pillars we have already outlined in the article: Brand; Content; Owned Media; Earned Media; Paid Media and GTM. It is the alignment across multiple functions and disciplines that demands a strategic approach. They all need to be guided, unified and measured by GEO analytics with overarching responsibility sat with Growth Leaders.

Done well, organisations will experience an increase in inbound demand which in turn can lead to significant GTM cost savings. We will explore this further in Part 3 of this article series but needless to say this change requires a mindset shift as much as any tangible strategic or tactical alterations meaning the transition can be hard, particularly in B2B that historically has been wedded to an outbound motion.

GEO implementation and LLM adoption supporting buying journeys are here to stay. We are in the AI era of buyer purchasing and vendor GTM. This requires fast adaptation by Marketing, Sales and GTM teams. The concept of RevOps (the unification of Sales, Marketing and Client Success inside a single operational framework to drive revenue) is gaining traction but it really is the tip of the iceberg when it comes to the GTM transformation required to embrace this new era of AI-led customer buying.

Look Out for the Next Article from this Author

Part 3: "The AI era of Customer Buying and the Future of Growth"

If you missed Part 1, check it out here: The Rise of the Machines: Why LLMs have become our Trusted Buying Companions | lomo ai