# GEO: Generative Engine Optimization

**Authors:** Pranjal Aggarwal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan, Ameet Deshpande  
**arXiv:** [2311.09735](https://arxiv.org/abs/2311.09735)  
**Date:** 2023-11-16  
**Summary page:** https://www.summarizepaper.com/en/paper/2311.09735v3/

## Key points

- Large language models (LLMs) have transformed search engines, leading to the rise of generative engines (GEs) for personalized responses.
- GEs are more efficient and relevant than traditional search engines like Google and Bing.
- Content creators struggle with ensuring their content is featured prominently in GE responses.
- GEO has been introduced as a novel paradigm to enhance content visibility in generative engine results.
- GEO employs a black-box optimization framework to define visibility metrics for website content.
- Implementing GEO strategies can improve content visibility by up to 40% in generative engine responses.
- Position-adjusted word count metrics consider citation position's impact on user attention.
- Subjective impression metrics evaluate relevance, influence, uniqueness, and other subjective aspects of citations on user engagement.
- Various methods help content creators optimize website content for better visibility in generative engine responses.
- Certain GEO methods outperform traditional SEO techniques like keyword stuffing according to evaluations using GEO-bench benchmark.

## Summary

The emergence of large language models (LLMs) has revolutionized search engines, leading to the development of generative engines (GEs) that utilize these models to provide accurate and personalized responses to user queries. These GEs have quickly gained popularity, outpacing traditional search engines like Google and Bing in terms of efficiency and relevance. However, while users benefit from this advanced technology, content creators face challenges in ensuring their content is prominently featured in generative engine responses. To address this issue,  has been introduced as a novel paradigm to assist content creators in enhancing the visibility of their content in generative engine results. GEO employs a flexible black-box optimization framework to optimize and define visibility metrics for website content. By implementing GEO strategies, content creators can improve their content's visibility by up to 40% in generative engine responses. One key aspect of GEO is the introduction of position-adjusted word count metrics, which take into account the position of citations within a response to determine their impact on user attention. Additionally, subjective impression metrics are proposed to evaluate the relevance, influence, uniqueness, and other subjective aspects of citations on user engagement. Various  have been developed to help content creators enhance their website content for better visibility in generative engine responses. These methods range from simple stylistic alterations to incorporating new structured content on websites. Through rigorous evaluation using a large-scale benchmark called GEO-bench, it has been demonstrated that certain GEO methods significantly outperform traditional SEO techniques like keyword stuffing. Overall, represents a significant advancement in information discovery systems, offering valuable insights for both developers of generative engines and content creators. By leveraging GEO strategies effectively, website owners can improve their visibility and ensure that their content remains competitive in the evolving landscape of search engine technology.