The origins of SEO and what they mean for GEO and AIO

The debate on SEO‘s changing practice and its transition to AI has heated up recently on podcasts, blogs, news sites, and social spaces around the web. 

While the discussion is focused on what we should call it and why – be it “GEO” (generative engine optimization), “AIO” (artificial intelligence optimization), or something else – one linguistic element keeps surfacing.

No matter the acronym, it will most likely include the word “optimization.” 

Most people debating the term likely do not know the details of its origins, as a similar debate about optimization occurred almost 30 years ago – before many of today’s debaters were even born.

While naming and debating the linguistics of this new thing may seem trivial at a high level, the fact is that the right time is now for the discussion to take place, just as the progenitors of SEO had these discussions back in 1995-1996.

Why optimization still matters

While AEO, AIO, and GEO are acronyms that have been bandied about a lot, many people seem to be vying to be “the one” to coin this new term. 

In the early days, there were literally tens of people doing SEO. 

As it progressed to hundreds, then thousands, it was still a small enough group where consensus could be reached.

Now, with millions engaged in the practice, don’t be surprised if current practitioners never agree. 

Many high-profile SEOs are now pushing into variations of GEO and AIO. 

Tim Sanders, Harvard fellow and SVP of AI evangelism at B2B directory G2, told me they had their own naming deliberation a couple of weeks ago, and they changed the G2 category listing from GEO to AEO.

But I would make the case that no matter what the new term ends up being named, if the moniker includes the word optimization in any form, then the credit goes to the original people who gave SEO its name.

The grammar problem with ‘optimization’

As practicing search marketers, we all know about the taxonomies and linguistics of keywords. 

The single word “optimization” in SEO identifies a stemmed property of a search engine, or in other words, “optimizing for” a search engine.

However, it is fair to say that the term still has a grammatical challenge, and this discussion around optimization for AI is not immune to the same problem.

Veteran search executive and pioneer Mike Grehan has posed the following question over and over at conferences and in columns and blog posts for more than 25 years: 

• “How does one optimize a search engine? You can’t.”

He’s right – when construed that way, the term SEO does not make grammatical sense. 

I have posited to him a few times that it can also be “optimizing for,” as the original creators intended. 

But the new naming risks the same grammatical issue with GEO and AIO.

---

Who coined SEO?

Back around 1995–1997, five people came to the “optimization” realization at the same time, though each was unknown to the others. 

Bruce Clay invented the term and has been SERP-famous for ranking first in Google for “search engine optimization” for over a decade.

John Audette and Bob Heyman (and his business partner Leland Harden) are also credited with coining the term, though their names are not included in the Wikipedia entry for search engine optimization.

Danny Sullivan should also get credit for helping popularize the term through his extensive search engine news coverage at Search Engine Watch and Search Engine Land. 

And in my interview with Heyman for this article, one other new name came up: Viktor Grant.

None of the people mentioned thus far has ever disputed – either in person or in print – that the others did not independently come up with the phrase. 

I spoke with four of the five for this article, and none of them disputes the others’ epiphanies as copies.

In my interview with Heyman and Grant, they also stressed that Stephen Mahaney of SearchEngineNews.com and Planet Ocean made key contributions in popularizing SEO, and it is true that he has largely been uncredited up to now. 

Grant also prefers to be known as a pioneer in black hat SEO, before it even had a name, as he was doing IP/agent targeting – later referred to as cloaking, among other names.

Clay said his use of SEO came from his previous work in Silicon Valley doing mainframe optimization. 

“Optimization” was already embedded in his mind when it came time to give this new concept a name. 

Speaking on the various new terms, Clay joked:

• “I have started to use the Old McDonald framework where it’s E, I, E, I, O.” 

But seriously, he continued:

• “The fact is, anything can be an engine, but technically, there’s nobody there saying it’s only organic. Technically, pay-per-click is SEO. But when you look at SEO, I think we as an industry have accepted that it’s on-page, off-page. And when you think about off-page and think about linking, those are practice areas within SEO. That is where GEO belongs, AEO belongs. Those are specialties within SEO under the umbrella of SEO.”

Heyman and Grant are proponents of the term GEO, and their opinions should carry significant weight in the overall conversation. 

Heyman said:

• “In terms of a marketing world… influencers have adopted it, so I think it’s going to catch on. And also because it plays off ‘SEO,’ [which] people are used to.”

The naming of SEO took much deliberation, as they were trying to solve a navigation issue with the build of a website for the band Jefferson Starship. 

Regarding the genesis of the practice, Grant said:

• “The concept of the 3 a.m. call from the Jefferson Starship manager is really the thing that cemented it. And it’s like, what do we call it? And Heyman looked at him and said, ‘search engine optimization.’”

For Jefferson Starship, no gaming of the system was involved. 

The fact is that the search engines of that time could not find the official band website for exact searches of their name. 

And those who were solely concerned with search issues at that time likely numbered in the hundreds or low thousands, as opposed to now, when millions participate or lurk in the conversations.

How SEO spread

Those early deliberations are also notable because optimization wasn’t exactly a slam-dunk term at the time. 

Frederick Marckini, founder of the first SEO-only agency iProspect (disclosure: I was a previous employee at this company), wrote a massive tech book in 1999 titled “Search Engine Positioning,” which was effectively the first major print book on SEO. 

Grehan also wrote one of the most significant books on the topic at that time.

For whatever reason, “positioning” didn’t stick, though it was a valid candidate.

If it had, how would the terms generative engine positioning and artificial intelligence positioning sound to you? 

Probably the same way AIO and GEO sound to me.

From GEO to today: Why names don’t stick

Here’s another issue with GEO. 

I’ve been talking about AI in one form or another for over 25 years.

Not once in my experience has anyone ever called an AI or LLM a generative engine. 

In that sense, it is largely a made-up phrase. 

What do we typically call it? 

• An LLM. 
• AI. 
• GPT. 
• A search engine. 

But never “generative engine.” 

For some reason, LLMO and GPTO just don’t seem to have the same ring to them.

Lastly, aside from the obvious semantic log-splitting of the term GEO related to geography in search marketing and everywhere else on the planet, there is another big issue with GEO. 

It paves over one of the biggest advantages of AI to businesses: agentic workflow and autonomous delivery – an element just as important to future success as being visible and found.

LLMs are like black magic for anyone who has ever written content on their own, or created art, images, video, in-depth research, or code on their own. 

Similarly, one of their most important uses is automation for a multitude of tasks in the search business. 

If you are still using LLMs but are still copy-pasting a lot, you are missing out on some of the main benefits and imperatives of AI in the modern world. 

GEO says nothing to address this key element.

I personally don’t think any one term will stick in the not-so-distant future. 

We are all in our own algorithmic filter bubbles, and we are all too tribal. 

Be aware that the different terms all have the same intention, and either correct or roll with it accordingly.

We may each end up speaking a different language, but the meaning is still the same.