New online search behaviour in the age of AI

Nearly half of the people who use artificial intelligence do so for online searches. Proof enough that AI is increasingly present in our lives. Searching for information online has always been a major and constantly evolving ecosystem. Fast-forward to today, and these new search behaviours are impacting companies’ online presence, putting two strategies front and centre: SEO and GEO. Read on to learn all about this shifting online landscape!

GEO is revolutionising the way we search online

Google has largely dominated the online search market for more than twenty years. This internet behemoth has continued to improve its search results pages to offer users ever more relevant answers to their queries. However, the arrival of AI and large language models (LLMs) has redefined the online search world.

Previously, users just entered short, precise keywords when using a search engine, for example “best sports watch”. Brand-wise, the aim was to use these keywords to guarantee a top slot on search results.

However, LLM logic is a whole new playing field and we’re now dealing with conversational searches: internet users ask longer, naturally-worded questions that refer to a context. Apply this to the above example, and you’ll get something like: “best sports watch for a marathon, nearby shop”.

This metamorphosis is radically raising the bar for companies’ online presence. It’s no longer just about getting clicks; companies need to ensure they are a trusted source to be cited in a conversational agent’s reply.

Example of a Google search plus a ChatGPT search that also displays a local branch

The different types of search in the GEO era

Traditionally, SEO search intents were grouped into five main categories that corresponded to a specific moment in the customer journey:

• informational (a search for knowledge),
• navigational (find a specific site or brand),
• sales (comparison before making a purchase),
• transactional (the user plans to make a purchase),
• local (looking for a nearby business or service).

However, the arrival of AI and conversational searches has completely reshaped this approach.

Today, only 30% of searches using conversational AI or conversational agents follow these specific intents. The remaining 70% are new “chattier” search intent behaviours centred on:

• troubleshooting: For example, an AI-generated reply to “how to glue the sole of a leather boot”, may suggest a list of suitable glues, where to buy them, or even a list of highly-reviewed local cobblers.
• seeking ideas: Ask an AI agent “what to do in Brighton during a summer holiday” and it may list tourist activities and relevant local businesses (surf schools, popular restaurants, etc.)
• image, sound or video generation.

In other words, the search is no longer limited to “finding information”, but to “converse with and work alongside” AI to get a personalise answer.

How can local companies or branches create a GEO-friendly online marketing strategy?

A recent study by Eskimoz shows that the companies highest up in local pack and Google Map results are also the ones most frequently cited in LLM-generated search results. This proves that local visibility remains a non-negotiable strategic lever, even in AI-dominant times.

In addition to core actions to drive your online presence (ensuring your Google My Business listing is complete and boosting your e-reputation via customer reviews), you must now gear your strategy towards GEO.

Harnessing the power of GEO requires building rock-solid foundations.

If you want a conversational agent such as ChatGPT, Gemini or Copilot to cite your company, you must present it as the most reliable and relevant source of information.

This requires producing content that:

• gives users a precise answer to their question,
• promises that visiting your website provides real added value,
• uses a clear and hierarchical structure, so LLMs can easily extract information.

The importance of structured data in GEO

Structured data plays a pivotal role in GEO. Think of it as “invisible tags” that ensure AI crawlers reliably interpret every element of a website.

Indeed, when an LLM generates a local response, it retrieves the structured data directly from the websites it crawls. In 60 to 70% of cases, this structured data comes from store locator pages. This interactive map lists your various establishments or branches, so internet users can find the nearest one, making it a goldmine of information for AI-generated search engines.

Finally, conversational agents prefer content written in a natural and understandable tone of voice. To maximize the likelihood of appearing in answers, we recommend you:

• adopt a conversational style,
• anticipate internet users’ questions and provide answers in your content,
• avoid sentences that are too wordy or technical.

Your content, therefore, needs to “speak” to humans and artificial intelligence if you want to capitalise on this new search behaviour and strategically exploit GEO!