For publishers and writers, the rules of SEO are changing. Future success no longer looks like being Top Source Media marketing of the blue links on Google’s index or any other search engine’s – it will center on how to ensure your content gets surfaced in AI answer engines too. And that’s a whole new ball game. Welcome to the era of generative engine optimization (GEO) – or is it answer engine optimization (AEO)? Actually, it’s generative search optimization (GSO), right? WTF is GEO and is it the same as AEO? This is such a new area, that currently there is no common taxonomy. So agencies, publishers, marketers and SEO specialists have adopted a bunch of different acronyms to describe the same trend – AEO, GEO and GSO for starters – and they all mean the same thing. Some are looking even further ahead – at agentic engine optimization. But that’s for another day, and another article.
In a nutshell, what all three of these acronyms represent is how all brands – whether you’re a publisher, a marketer, an e-commerce site – ensure that AI crawlers can easily understand enough information about your brand in order to surface it in the synthesized answers they curate within their answer engines. If that’s part of your strategy. “AI has already radically changed what search means, not just in terms of the look and feel perspective, but also in terms of what search is, the kinds of queries or asks,” said Tom Critchlow, evp of audience growth at ad management and creator monetization platform Raptive. Will it replace SEO then? Hell no, search is still far too ubiquitous and deep-rooted, but this is being developed alongside it and is being regarded as a strategic imperative by SEO experts. But there is no standardization around it yet, and certainly no playbook.
No doubt, that will change fast. So how does it compare with SEO? Traditional SEO has centered on finding ways to make your website appear high up within the search indexes when people are looking for specific things online, measured with CTRs. GEO/AEO/GSO is more about ensuring the AI crawlers have enough information to be able to answer a potentially more complex question written by the user, within an AI engine itself. “GSO is really more about understanding what is happening within that AI result, the response that’s been generated from your prompt, how your brand is represented there, whether or not it’s linking back, whether or not it’s citing your content or your brand correctly, and whether the information is accurate,” said Cowell. What do AI bot crawlers look for, and how does that compare to traditional search crawlers? In comparison to search engine crawlers, the large language models and their crawlers are still “pretty crude,” given the head start search engines have had, said Cowell.
That means they do have issues around being able to access all the content from some websites. An option getting talked about is just how far a brand should go in making its website content simpler and easier for AI engines to pick up. Currently, you can, for example, upload all the raw data from a website into LLMS.txt to make it available to the LLM – a bit like the flip side of robots.txt. How do you optimize for AI engines versus search engines? Much of it comes down to the fact that the kinds of questions users are putting into AI engines like ChatGPT are radically different from what they would typically type into search. That’s what some SEO buffs now refer to as “long-tail queries” – lengthy prompts that an individual writes into an AI engine, which delivers a synthesized answer, instead of a more simplified, keyword-based question that a user would typically write into a search engine to deliver a site to click through to.
’re very hard to predict,” said Mollie Ellerton, head of SEO at digital optimization agency Hookflash. “The difficulty, at the moment, is that there aren’t any Webmaster tools that report on performance within AI overviews. So Google gives us no data on AI overviews. Similarly, we don’t get reports that we can dig into from ChatGPT, for example. So already we’re a little bit blind,” she added. Are there any workarounds? Some yes, though they’re likely not 100% accurate. “A quicker win, especially if you’re a content marketer, if you’re writing a piece on a specific topic, is to go to Reddit and look at what people are saying around that topic,” said Ellerton. ” She believes that a content creator can then have a deeper understanding of how to tailor their content to the types of questions people are asking on these platforms, and possibly have a better chance of surfacing prominently in an AI engine as a result.
