Certitude has long been a nemesis of mine.
Since my life has largely involved the internet, that can be hard. As a medium, the web loves strong, declarative positions, and abhors restraint of opinion and its close cousin, nuance.
In the GEO world that so occupies my and my team’s brains these days, certitude usually means someone is either ill-informed or hasn’t been paying attention at all.
Because if anything is certain, it’s that the field is moving too quickly for certainty.
That said, over the past 14 months, and thousands of hours, we have developed a perspective, we believe in right now. Some views mirror conventional wisdom. Some don’t. Parts are based on larger data, parts based on very anecdotal evidence and small samples of experience, some are based on intuition and inference (crucify me if you must, data nerds).
Could our views change in 2 months or 6 months? Of course.
But, right now, at this point on the space / time continuum, here are the core things Proper Propaganda believes about GEO:
I’ll start with the obvious and vanilla. The bigger picture is pretty well undeniable based on good data and our micro-level experience:
- AI is the new thing between sellers and buyers – for both B2C and B2B. Usage stats make a strong case about the importance of, and the why behind, GEO. AI Overviews now reach more than 2.5 billion monthly users. ChatGPT sits at 900 million weekly active users. AI is how buyers find and decide things. Bots are every organization’s new ICP.
- AI is present across the entire buying journey with a buyer. I have talked about AI as the passenger in the motorcycle sidecar. Buyers of considered purchases (not those making routine or impulse buys) consult it from discovery, to evaluation, to purchase and into post purchase. Our thinking has been shaped by this very interesting data set, as well as our own experience running our system on ourselves and clients. AI is still viewed a neutral oracle by many Americans, it appears. That will likely change, but for now it’s a defensible point of view.
Our perspective on the strategic and executional components of GEO programs runs from the totally conventional to “I can’t believe no one talks about this.”
- Because each LLM varies in terms of what it cites, people who pursue GEO need to treat each as a separate entity. This is entirely in line with conventional wisdom and the broader convo. The data is clear on this and has been for 3 consecutive quarters. It is a fact, for now. How you optimize for one platform differs from how you optimize others.
- You should not start a GEO program before you know what LLM(s) your customers use and what prompts they enter. Unbelievably, all the big GEO brains on the internet fail to mention this with any consistency. Maybe it is so fucking obvious and we are the idiots. But it HAS to be an early stage move in any program, lest you end up optimizing for the wrong LLM and giving your cash a death by a thousand paper cuts. You must survey to understand these things. Failure to do so means you are guessing.
- Aggregate data such as Muckrack’s “What AI is reading” is useful but individual client data is where the real gold is. So much of the conversation about GEO revolves around people making generalizations based on aggregate data. While Gemini may indeed cite Reddit heavily and ChatGPT may cite Wikipedia, a company’s own AI search program often shows completely different citation patterns. Our most cited site over 700 prompts is our own ffs. Our consumer tech clients often see weird, tiny websites no one ever heard of being the main places AI goes to to formulate answers.
- GEO involves disparate areas, and at least web dev, content, PR and SEO. You need PR in a powerful position. Hardly a revolutionary perspective. But, because 27 percent of AI citations are from journalistic sources, you have to have PR as the lead dog on your multi-disciplinary GEO team. Are we biased? Hell yes.
- Be careful who you trust for info, even if their name is Google. As I said earlier, there is a pile of advice out there that’s weak or too generalized. Google, who you might think had enough data to make reasonable pronouncements, has regularly said things that run counter to what we see for ourselves and clients. In particular, the tech giant loves to exhort everyone to write content for humans, by humans with human soul. Well, a range of content written by this human, with what I’d argue is nice prose and pizzazz, has been de-indexed, likely because it was too opinionated and not fact-stuffed GEO fodder. Likewise, some of the best performing content in our GEO program, is bot written for a bot audience. Go figure. Your data and your experience are the best guides for your GEO program, not platitudes from tech kings on high.
I began this piece talking about my issues with certitude. This final point, we are pretty certain about.
- We are in a moment in time, one where smaller players can be bigger ones at GEO. This applies to B2C and in B2B. The risk of funding and trying your hand at a relatively immature channel is greatly outweighed by the upside available to early movers and huge risks associated with “waiting till this matures.”
There are many other things we think about GEO. I want to call that out as being distinct from the above eight points and instead beliefs that are held less strongly.
I’ve outlined our core perspective. And the basis of our programs. Pillars that hold now, but could very well change.
If you are interested in working with us, these are important things to know about.
Do you have core beliefs about the topic? I’d love to hear them.