Will it be RIP SEO Soon? The Dawn of GEO: Optimizing for the AI Search Revolution

4 min readJan 29, 2024

I am a smug man today. Why? One of the many forecasts I made following the introduction of ChatGPT and the other LLMs is turning out to be true.

Way back in 2019, and then more recently in February and May 2023, I wrote that gen-AI would be a pivotal moment for search engine optimization (SEO).

Here’s what I wrote in 2019:

Search algorithms are still evolving. Till such time that algorithms driven by artificial intelligence and natural language processing do not become “brainy enough” to comprehend intent, such optimization specialists have to be around. Once they are able to, I predict SEO shall be done and dusted.

Then again, I said this in early 2023:

Together, these two technologies (gen-AI and AI Co-pilots) have the potential to create a new kind of search that is more powerful and efficient than anything we have today.
I have been one of the few to have stuck their neck out and forecasted that both “search engines” and SEO will die. Eventually.

I stand vindicated on one count at least. SEO may have to give way to a new concept called Generative Engine Optimization (GEO) for AI search engines, which are advanced systems like Google Bard and BingChat that provide comprehensive responses.

A new study by researchers from Princeton, Georgia Tech, The Allen Institute for AI, and IIT Delhi has shown that traditional SEO methods may not work well with these gen-AI engines. Which made the team evaluate strategies to improve website visibility.

What is GEO?

GEO proposes a paradigm shift in how we optimize online content. Unlike SEO’s focus on ranking in traditional search engine results pages (SERPs), GEO aims to increase a website’s visibility in the responses generated by AI-powered engines like Google’s Bard or Bing’s LaMDA.These engines rely on large language models (LLMs) to synthesize information from various sources and provide comprehensive, tailored answers to user queries.

The study compared classic search engines with generative engines. It outlined nine GEO methods, such as including citations and quotations. The most effective methods were Cite Sources, Quotation Addition, and Statistics Addition, which improved visibility by 30–40%. However, some strategies, like using persuasive language or keyword stuffing, were less effective.

How is GEO Different From SEO?

The game changes when it comes to understanding how LLMs process information.Think less about keyword stuffing and more about semantic richness, factual accuracy, and structured data that enables engines to comprehend your content efficiently.Additionally, user intent takes center stage, as you optimize for the context and nuances of potential answers, not just generic keywords.

The Research Behind GEO

The paper, “GEO: Generative Engine Optimization” proposes a framework for optimizing content for generative engines. It introduces GEO-bench, a benchmark of diverse queries across multiple domains, to facilitate systematic evaluation. Their research shows that incorporating relevant statistics, quotations, and citations can significantly boost content visibility in generative engine responses.

And…here’s some bad news for the present lot of SEO experts: the study revealed that the keyword optimization strategy, which entails incorporating more keywords from the search query into the content, did not yield favorable results for GEO. In fact, keyword optimization performed 10% worse than the baseline. Consequently, according to the findings in the article, keywords are not effective for GEO.

The primary findings of the study showed that generative engines require a distinct set of impression metrics to assess the visibility of citations and their relevance to user queries.

The researchers established a benchmark using data from nine distinct sources, encompassing 10,000 search queries spanning various knowledge domains and levels of complexity. The top three successful optimization strategies identified were:

Source Citation
Inclusion of Quotations
Addition of Statistics

These three strategies demonstrated relative improvements of 30–40% compared to the baselines.

Additionally, the study emphasized the crucial role of domain-specific optimization strategies, as various GEO methods demonstrate superior performance in specific domains.

The team, by using various methods, effectively boosted the visibility of smaller, lower-ranked websites by 115%, granting them the capability to surpass larger corporate sites that typically dominated the upper echelons of search results.

Will GEO Replace SEO?

It’s still early days, but the researchers envision a future where generative engines become the primary gateway to information. However, complete SEO obsolescence seems unlikely, for now. Traditional search engines will likely still serve specific needs, and SEO skills can be adapted to GEO principles. Think of it as an evolution, not a revolution. For some years at least, SEO will learn to co-exist with GEO, till online search and the web themselves become obsolete.

What Does This Mean for Us?

The landscape is complex, but the takeaway is clear: staying ahead of the curve in the AI age requires embracing new concepts like GEO. We need to adapt our content strategies to cater to the nuanced information processing of LLMs and understand how users will interact with generative engines. This is an exciting opportunity for content creators and SEOs to evolve their skillsets and play a pivotal role in shaping the future of search.

Remember, this is just the beginning of the GEO journey. As research progresses and AI technology advances, the optimization landscape will undoubtedly continue to morph. So, keep your learning hat on, and embrace the AI search revolution with open arms (and minds)!