Andrew Higgins is CEO of Parsnipp, an AI search marketing company.
Everyone is talking about generative engine optimization right now. There is a lot of buzz, a growing class of experts and gurus and no shortage of companies positioning GEO as the new way marketers need to think about discovery.
My concern is less about GEO than about the certainty with which its playbook is being sold.
We are very much in the first inning. The category is maybe two years old. It changes every few weeks. Consumer behavior is still forming, the labs are constantly shipping changes and much of what marketers are doing remains untested. Yet we are already talking about the discipline as though we have a settled playbook.
Any strategy being positioned as the answer may eventually prove correct. But right now, it is still untested. The frenzy is misplaced.
GEO Has A Measurement Problem
One of the biggest problems is that even the underlying data is not clean or clear yet.
Marketers are accustomed to analytics on our websites, Google Search Console and keyword trackers. With AI search, much of the industry is effectively simulating the data because we cannot see exactly how people are interacting with large language models.
That is not inherently bad. It just needs to be more clearly acknowledged.
If step one of any marketing exercise is understanding the data and measuring it, we have not fully established step one. Talk to 10 GEO experts and ask what to measure and how to measure it, and you may get 10 different answers.
That makes it difficult to credibly say we have fully baked solutions for how brands should reposition their content, marketing strategy or investment.
There was a time when marketers could comfortably say, “We’re testing this because we don’t know which ad works.” Somewhere along the way, acknowledging uncertainty started sounding less sophisticated than claiming certainty.
With GEO, that posture needs to come back. Learn, test, measure what you can and stay nimble.
We Have Seen This Movie Before
Strip away the AI rhetoric, and this is following the same evolution as almost every new marketing channel. A new destination appears where consumers make real decisions. Search did it. Facebook did it. TikTok did it. Consumer behavior shifts quickly, and marketers follow.
Then come the hacks. Marketers hid keyword text against matching backgrounds so crawlers could read it while humans could not. We stuffed keywords into pages. Early Facebook apps grew by automatically spamming users’ friends. “Like this page to unlock the content” became a growth strategy.
Those tactics worked. Some worked extremely well. But they did not last because they were exploiting shortcomings in the system rather than helping the platform do what it ultimately wanted to do for users.
That is the test I would apply to GEO today. Does the tactic help an AI system better understand what you sell, who you serve and why you are relevant? Or are you simply exploiting the fact that the system is not very good yet?
It is perfectly fine to use a tactic that works today. What is dangerous is mistaking that tactic for a permanent marketing strategy.
The Real Change Is The Surface
The most important development here is bigger than GEO. The entire surface we interact on is changing.
Consumers once interacted directly with a business. Then the telephone became an intermediary. Then websites, search engines and social platforms became the surfaces through which people found and interacted with brands.
Now that surface is increasingly AI. People are asking an agent instead of going directly to a retailer, a brand website or a social feed. Suddenly there is another layer between the customer and the company.
For the last 30 years, our job was largely to get customers somewhere we controlled and convert them there. Now a middle layer can do the researching, the comparing and, increasingly, the deciding. The interaction is no longer necessarily with you.
That’s why I think GEO is being framed too narrowly. Generative engine optimization is essentially the organic slice of marketing to AI. It is important, but it is one slice. SEO was one piece of digital marketing. GEO will be one piece of AI marketing.
Advertising will matter. Commerce will matter. Brand will matter. And as agents consume more of the customer journey, marketers will have to think about all three together.
Do The Durable Work
None of this means brands should ignore GEO.
There are obvious things they should be doing now. I think of them as technical hygiene. Structure your website correctly, improve your product feeds, use emerging protocols that make products readable to agents and make your content easier for models to understand and cite. Most of that is simply good digital practice.
Where I would be far more cautious is changing the way a brand communicates in order to accommodate whatever an LLM happens to reward today. Flooding models with low-quality content is essentially the 2011 content-farm strategy with new paint. Mass-producing 75 derivative versions of the same article because an AI system might ingest them is not a long-term brand strategy.
Brand may actually become more important as AI gets smarter. If a model is synthesizing what the world says about you, you cannot shortcut reputation. You still need real brand development, good writers, reputable partners, strong PR and high-quality content.
Nike is Nike because of what the brand means, not because it can produce 9,000 generic blog posts engineered to answer prompts.
What Comes After GEO
GEO will remain relevant. What changes is everything around it.
AI search marketing will become more holistic. Organic will have to interact with advertising and agentic commerce because that is increasingly how the customer may experience the journey: ask an agent, get a recommendation, buy. One motion.
Today, the marketer’s tool belt is still thin. The available data, targeting and use cases are immature. That will change dramatically.
So instead of asking only, “What is our GEO strategy?” marketers should begin asking, “What is our approach to AI as a channel?” Do the boring, durable work now. Start measuring so you have a baseline. Test and learn. Stay genuinely flexible while the labs, algorithms and consumer behavior are all moving at once.
And be wary of anyone telling you they already know exactly where this lands. The current GEO playbook may change three months from now. The shift toward consumers interacting with agents will not.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
