GEO Strategy: How to gain brand visibility from 'AI Search'
- James Harbottle
- 5 days ago
- 6 min read
Updated: 4 days ago
AI Search has certainly become a topic of relevance in the marketing world, and indeed, everyday life.
According to Data Reportal, over one billion people use Ai search platforms, globally.
(Source: Data Reportal, 2025)
Interestingly, ChatGPT actually launched it's first iteration (GPT-1) in 2018, which was superseded, before GPT-3.5 was launched in November 2022. Google Gemini followed suit in March 2023; amongst other LLM's (Large Language Models) that quickly gained in popularity.
Then came the process of GEO, Generative Search Optimisation.
What is GEO and why does it even matter?
Generative Engine Optimisation is the process of optimising your brand / product / service for Generative Search in LLM’s.
Generative Search Engines include, but are not limited to, platforms such as:
· ChatGPT
· Claude
· Perplexity
· Gemini
Large language models don’t index the web like search engines, but they do rely on patterns (Consistency), associations, and widely referenced information found across high-quality sources.
LLM’s care most about what people are saying about you, not what you are saying about yourself, hence online sources showing that forums provide 40%+ of LLM’s intelligence.
(Source: Search Engine Land, 2025)
Your domain alone isn’t enough anymore. This was our biggest breakthrough. You'll need to optimise for mentions on Reddit, Quora, Medium, LinkedIn, industry forums—anywhere LLMs might discover your expertise.
(Source: SEMRush, 2025)
To change the subject slightly, If you’re familiar with off-page SEO, the process of establishing ‘follow’ links from trusted high authority websites, back to your site, you will understand the process of Generative Engine Optimisation is similar; however relies on brand mentions, positive brand affirmation and positive experiences of brand/services/products.
Primary Research
There are a lot of theories online as to where LLM’s source their information, so it seemed diligent to conduct some basic primary research.
We conducted research on ChatGPT & Gemini for mainstream and niche search queries to see if there was a difference.
First up was a very niche search query relating to boat engines.
Niche Search Query: Is the Evinrude 115hp ETEC outboard motor reliable?
Chat GPT (Free version) | |||
Niche Search Query: Is the Evinrude 115hp ETEC outboard motor reliable? | |||
Query Search Volume: 0-10 (Google Ads) | |||
Type | Source | Notes | SEM Rush Site AS |
Forum | No keyword, but partly present in URL | 12 | |
Blog | No keyword, but party present in URL | 2 | |
Blog | No keyword, but URL and H1 relevant | 5 | |
Google Gemini | |||
Niche Search Query: Is the Evinrude 115hp ETEC outboard motor reliable? | |||
Query Search Volume: 0-10 (Google Ads) | |||
Type | Source | Notes | SEM Rush Site AS |
Forum | No keyword, but URL and H1 relevant | 8 | |
Blog | No keyword, but URL and H1 relevant | 17 | |
For this very niche query, both Chat GPT and Gemini pulled data from fairly low authority blogs & forums with semi-relevant information in the URL and H1, as per the above ‘notes’ columns.
Next we tried asking both Chat GPT and Gemini recommendations for the best winter mountain bike clothing, considering the UK climate. Although niche, this is more of a main stream query.
Niche Search Query: Recommendations for the best winter mountain bike clothing, UK climate?
Chat GPT (Free) | |||
Niche Search Query: Recommendations for the best winter mountain bike clothing, UK climate? | |||
Query Search Volume: 0-10 (Google Ads) | |||
Type | Source | Notes | SEM Rush Site AS |
Blog | Relevant to the query, questionable source | 2 | |
Brand | Heading is very relevant, good content length | 60 | |
Digital Magazine | Heading is very relevant, good content length | 45 | |
Google Gemini | |||
Niche Search Query: Recommendations for the best winter mountain bike clothing, UK climate? | |||
Query Search Volume: 0-10 (Google Ads) | |||
Type | Source | Notes | SEM Rush Site AS |
Brand | Relevant title and H1, content also relevant and 1000+ words | 46 | |
Brand | Relevant title and H1, content also relevant and 800+ words | 49 | |
Blog | Relevant content, over 1000+ words | 32 | |
For the mid volume niche query, both Chat GPT and Gemini pulled data from higher authority blogs, forums & branded sources. It’s worth noting that the branded sources do have significantly high authority scores (SEM Rush) and these sources are well established brands.
Lastly, we tried asking both Chat GPT and Gemini recommendations for men’s running trainers. This is a main stream query that has high search volume in the UK.
Search Query: best men’s running trainers
Chat GPT (Free) | |||
Search Query: best men’s running trainers | |||
Query Search Volume: 100-100K (Google Ads) | |||
Type | Source | Notes | SEM Rush Site AS |
Review Platform | Relevant title & H1, 9000+ words, article schema | 67 | |
Digital Magazine | Relevant title & H1, 1700+ words, article schema | 17 | |
Digital Magazine | Relevant title & H1, 900+ words, News article schema | 72 | |
| Poor on-page SEO, 450 words, no schema | 79 | |
Interestingly, ChatGPT displayed some links to products from retailers, these products were also featured in the cited content.
(Image: Screen shot from Chat GPT)
Google Gemini | |||
Search Query: best men’s running trainers | |||
Query Search Volume: 100-100K (Google Ads) | |||
Type | Source | Notes | SEM Rush Site AS |
Digital Magazine – Review Article | Good on-page SEO, 6000+ words. In depth Schema. | 72 | |
Digital Magazine – Review Article | Ok on-page SEO, 5000+ words. Warning for Link Farm. In depth Schema. | 40 | |
Digital Magazine - | Average on-page SEO, 8000+ words. Schema. | 61 | |
Digital Magazine | Poor on-page SEO, 700 words. count. Schema present. | 37 | |
Digital Magazine | Good on-page SEO, 8000+ words. In depth Schema. | 45 | |
Digital Magazine | Present on-page SEO, 2000+ words. In depth Schema. | 14 | |
Digital Magazine | Good on-page SEO, 10,000+ words. In depth Schema. | 44 | |
For the high-volume search query, ChatGPT pulled information from review platforms and digital magazines with high authority scores. Chat GPT also provided links to product listings. Gemini did much the same, citing information from digital magazines, generally with high authority scores, however did not supply links to product listings unless requested.
Conclusion:
Our findings were that for low volume queries, both LLM’s cited information from low authority, often poorly optimised blogs and review based platforms, however for high volume queries, the information was cited from high authority (often 50+ AS) websites with good on-page SEO and structured markup (schema) language In place.
How do I go about getting my brand / products / services featured by LLM’s?
From the findings above, to increase the likelihood of a brand, product or service appearing in model responses, the priority is to strengthen presence within high authority digital magazine’s, blogs and forums.
Top Priorities:
News: | Your brand / product / service needs to be featured in a topical article. |
Platform: | Industry publications, Articles, Review Platforms or Digital Magazines with high relevance to your product of service. |
Performance: | These 3rd party site needs to have a high authority scores. |
Format: | The piece needs to be SEO optimised, high word count, in some cases 5000+, with correct use of schema. |
From our findings, this is the ‘ideal scenario’ and difficult/costly to achieve with or without the right connections.
In reality, for lower volume queries, we saw content featured from branded sites and also 3rd party sites with very low authority scores.
- Company websites
- Industry publications, blogs and forums
- Trusted directories
- Relevant third party mentions
It's far easier to get low search volume, long tail niche query answers featured than high search volume, short query answers, as you would expect. (Much like SEO).
On-Page, Off-Page & Technical SEO
Much of our primary data shows that good SEO principles help towards being featured in LLM’s, so it’s important that your On-Page, Off-Page, Local and Technical SEO is up to scratch, highlighting the fact that the majority of pages cited were using schema correctly.
GEO Strategy model
In essence, the path to visibility in LLMs mirrors the discipline of long-term content authority similiar to the Google EEAT principle: high-quality information, distributed widely, consistently, and credibly across high authority blogs, forums, digital magazines and also social media; with high relevance to your industry.
The below model represents in very basic terms, the channels that feed LLM's with citable information. (Taken from our primary research).
The process of Generative Engine Optimisation is about being mentioned and perceived as credible, allowing the user to do their research and make up their own mind, as opposed to 'clicks'. Clicks come potentially via direct traffic, branded search terms through traditional search, social media or even visiting a retail outlet.
Integrating GEO strategy into your current marketing efforts now, will ensure you are able to compete in the future, whilst improving brand visibility. Content planning and allocating responsibilities amongst the marketing team are key to GEO success.
Although it's possible using analytics tools to track acquisition from LLM's, it's far more useful to use third party tools such as SEMRush Prompt Tracking to measure visibility.
We hope this article has been useful. If you're looking to strengthen your GEO & SEO marketing, or if you have any questions, please don’t hesitate to contact-us. Disclaimer: The views and opinions expressed in this article are solely those of the author and do not necessarily reflect the official policy or position of any other agency, organisation, employer, or company. All references and data sources have been cited appropriately. The information provided is for educational purposes only; the author accepts no liability for any results or actions taken based on the findings presented in this post.