I just went through this article from @Roger Montti covering this paper from researchers from Princeton University, Georgia Tech, The Allen Institute for AI and IIT Delhi testing methods to optimize the visibility in AI results, introducing "Generative Engine Optimization" (GEO).
They tested 9 methods, with 3 getting improvements of 30-40% vs baselines": * Cite Sources * Quotation Addition * Statistics Addition
"Among other things, we find that including citations, quotations from relevant sources, and statistics can significantly boost source visibility. Further, we discover a dependence of the effectiveness of Generative Engine Optimization methods on the domain of the query."
Check it out: https://arxiv.org/pdf/2311.09735.pdf
Are you ready to GEO?
Do you Expect Generative Engine Optimization (GEO) To Become Part of SEO in the Near Future?
Yes
No
Their findings are quite interesting. I've been testing them out with SGE and Bing Chat, but have not had positive results as of yet.
Being an SEO, one of the things that stuck out was their misunderstanding of SEO as well as their downplaying of the role of SEO in AI visibility.
For example, they referred to using additional instances of keywords as "keyword stuffing." As we know, that in and of itself is not keyword stuffing. I am assuming they just meant keyword optimizations but mixed up the terminology.
The other aspect is that they were measuring the performance of the top 5 SERP results. Specifically how their strategies impacted visibility based on their SERP position. It seems to me that if the AI is sourcing from those top results, SEO is playing an important foundational role.
As far as generalizability, the researchers used ChatGPT 3.5 to test. This provides no insight into whether their optimizations would work in Google SGE, Bing Chat, Claude, or any other AI.
If you are interested, I wrote an in-depth critical review of generative engine optimization (I hope it's okay to share this).
I definitely recommend reading the original study that Aleyda linked above and the articles on Search Engine Land and Search Engine Journal as well.