
Generative Engine Optimization (GEO) is a new-ish term that seems to pick it up in popularity, whether we want that or not.
The first time I talked about it was in 2023 when I was sharing a study on how to get cited in AI-driven answers. Since then a lot has changed, and with our industry’s love for new shiny terms, it looks like we need to address the topic again.
>>> Join us to discuss this LIVE and ask your question!
What is GEO?
GEO is optimizing for generative AI.
In more words, it is a strategy aiming at increasing your site’s and your brand’s chances of being included in AI-driven answers.
Why is it more complicated than it sounds?
Where do I start?
First of all, the term “generative AI” is too generic. Even if we only talk about the most popular gen AI platforms, there are key differences in how they work:
Google’s AI Overviews summarize high-ranking URLs
ChatGPT and Gemini rely on Bing’s and Google’s indices (respectively), but not so much on rankings
Deepseek, Claudia, etc. mostly rely on their current knowledge bases/training data (and these are usually quite outdated)
Perplexity is supposed to have its index (but it also scrapes Google).
Consequently, the optimization strategy is supposed to be different, but in reality, it is still good old fundamentals (content, links, tech):
If you don’t trust me, listen to Bing: Bing’s Advice on Optimizing for AI Search Engines. Everything is New Here </sarcasm>
GEO “experts” will tell you there are indeed core differences
The huge argument of new GEO experts is that you can and should actually train AI platforms by creating content explaining what your brand does.
We’ve talked a lot about this previously, referring to that as brand knowledge content.
What I fail to see here is how it is a new tactic. It is still content. It's still branding.
It may be a bit different tactic now, because you should provide more brand knowledge content these days than you would have previously.
But it’s not enough to warrant the whole new term.
In essence, it’s the same core principle: We (SEOs) reconcile the needs of machines and humans, whether it’s a traditional search or a gen AI platform.
SEO is likely becoming an outdated term
… because we are now optimizing for more than organic search (and we have been for a few years).
Ever since Google’s search results have become more than ten blue links, SEO has been more than “organic search optimization.” And now it is also more than “search” optimization.
But the strategy, at its core, remains the same: Making a site and a brand easy for the machine to understand to improve that site’s and brand’s findability.
Someone said on LI today, "how long till we just call it 'digital optimization'?" good question.
Ann, you are a staunch defender of traditional SEO despite outcries against it, and I appreciate that. It's daunting enough to use good old SEO strategies and tactics as they are, much less add an entirely new set geared to AI.
I especially appreciate your reference to brand knowledge content. I'm going to study that in-depth.
One question, though: Does getting favorable mentions in genAI mean that the machines will scrape our content? If so, does it work like the Google spider? That's something I'm confused about that I hope you can help me with.
Thanks for sharing your expertise!