← back
What is GEO?

What is GEO?

Generative Engine Optimisation (GEO) is a term used to describe SEO for AI assistants and generative search engines, often based on a single research paper.

GEO stands for Generative Engine Optimisation, an acronym easily confused with, the well-established “geo-” prefix commonly associated with Geosciences.

What is a ‘Generative Engine’?

Generative engine is recently made up term by the marketing community in an attempt to rename Chatbots, more recently known as AI Assistants including ChatGPT, Claude, Grok, Gemini and Perplexity.

Basically “Generative Engine” is meant to be a “Search Engine” powered by a generative model. A place where you search for information and get answers in a conversational manner, where the output is not a list of results but a generated text.

But there are already established names for all that.

Models: GPT-4.5, Llama 4, Gemini 2.5, Grok 3
AI Assistants*: ChatGPT, Claude, Gemini, Grok, Meta AI
Search Engines: Google, Bing, Yandex, Baidu, Naver

What’s new?

1. Search Engines can now be augmented by AI Assistants (AI Mode).
2. AI Assistants can now be augmented by search engine results (RAG) and allowed tool use via function calling and code execution.

What’s coming?

AI Agents. AI Assistants which agentic capabilities (Operator and Manus).

*Formerly “chatbots”.

The Origins of ‘Generative Engine’

It’s wobbly. A recently deleted Wikipedia article has been providing the term GEO most of its legitimacy on the basis of a 2023 preprint “by researchers Gao, Liu, Si, Meng, Xiong, and Lin”.

A screenshot of a Wikipedia article about Generative engine optimization shows a reference tooltip with incorrect author names.

But if you look it up Gao, Liu, Si, Meng, Xiong, and Lin don’t appear to be associated with this paper at all. Instead we’re looking at an Indian team of authors Pranjal Aggarwal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan and Ameet Deshpande.

The rest of the Wikipedia article are simply promotional, commercial links to various SEO blogs, providers and tools. The references are cyclical and loopy in nature as all these blogs borrow the term from this supposed Gao et al. paper.

Here’s another sneaky tactic used to attempt to legitimize the term:

A screenshot of a Wikipedia article explaining generative engine optimization and referencing researchers Gao, Liu, Si, Meng, Xiong, and Lin.

Notice how the word “generative engines” links to another article?

Yeah, it just links to Generative AI which uses the word “engines” precisely twice and to describe “search engines”.

A screenshot of a Wikipedia article showing highlighted search results for the word engines in two different paragraphs.

It doesn’t mention GE / generative engines at all. The wikipedia article continues to introduce more terms to describe SEO in the context of AI via a term LLMO with a citation to “Vibe Central”:

A Wikipedia article screenshot showing text about generative engine optimization with a pop-up reference citation for VibeCentral.

In summary, we have one research paper authored by Pranjal Aggarwal from Indian Institute of Technology Delhi dated 28 Jun 2024 and not 2023, Gao et al.

Aggarwal, P., Murahari, V., Rajpurohit, T., Kalyan, A., Narasimhan, K., & Deshpande, A. (2024). GEO: Generative Engine Optimization. arXiv preprint. https://arxiv.org/abs/2311.09735

Paper, code, data: https://generative-engines.com/GEO

It is my view that a single paper, especially when surrounded by self-referential citations and unsupported claims, does not justify dedicating a whole Wikipedia article to it pretending to be a whole new industry.

So many people are gonna be angry about this one. Especially those of y’all who’ve been trying to make AIO, AEO, LLMO, GEO, or EIEIO take off. But, I just couldn’t stand by and ignore the fact that the marketing world already has a perfectly reasonable, even excellent, acronym for getting visibility in all the places your audience pays attention… SEO.

Rand Fishkin

Dan Petrovic · Jun 02, 07:45