Recently, I thought of writing about all the good things that came out of my decade-plus career working with Search Engine Optimization (SEO.) I often focus on the ill effects when I write about it today, but there are things, especially when it comes to hyper-local and long tail keyword research that are helpful in my post-digital marketing career life still. 

On the same night, I asked myself a pretty big question: how do I break the metagame of anything? I don’t just mean competitive Pokemon or trading card games or video game speedruns. I believe that any field of study or pursuit can be boiled down into a “meta” that can be relentlessly studied in how people have optimized their own approaches. But to really “break” the metagame, in this case overcoming the average-seeking nature of SEO, you have to consider what may have been overlooked. 

So, this is all a really long way of asking, how do we break the metagame of SEO to entirely focus on the long tail keywords that are focused solely on spontaneous curiosity, while simultaneously taking a “hyper local” approach on figuring out how to reach other people on perpendicular vectors to your own curiosity habits?

That’s also a really long way of asking, how do we break the metagame of the SEO world, and stimulate our curiosity?

A few things I discussed with Gemini in the course of creating this prompt.

“SEO is essentially just the anthropology of human desire and taxonomy,” Gemini mused. “The current SEO meta is entirely built on homogenization: find the highest volume of searchers, match the most obvious “intent,” and regurgitate a slightly longer version of what already ranks. It optimizes for the average.”

What we want to produce then is intentional inefficiency. “Spontaneous curiosity is inherently inefficient, wildly messy, and deeply weird,” Gemini said. So, how do we create an anti-meta SEO framework that allows us to focus on serious questions that someone may not even think to ask directly, but may be thinking about along some curiosity vectors perpendicular to your own?

First, we need to generate potential long-tail search queries for our subject, which can be either broad, highly specific, or anywhere in between.

“The tools everyone uses to play the SEO game (Ahrefs, Semrush) scrub out anomalies to present clean data,” Gemini explains.“They filter out the 3:00 AM existential questions because only three people a year ask them. To capture spontaneous curiosity, you have to optimize for the anomalies.

“Instead of targeting a broad, highly contested term,” Gemini suggests, “you target the full, unedited sentence that crosses a beautifully unhinged mind. Don’t write for ‘classical literature analysis.’ Write for, ‘Why does reading classical literary criticism make me feel profoundly alienated from modern institutional education?’”

Honestly that’s a pretty solid question.

“The people typing those hyper-specific, long-tail phrases aren’t looking for a listicle; they are looking for a lifeline.” Gemini offered. I’m stealing that line. “When they find you, the connection isn’t a transactional bounce [which is what I’m used to]—it’s immediate intellectual kinship.”

Second, we needed to iron out the rhetoric of what I meant by perpendicular vectors. This whole exercise is based somewhat on Henrik Karlsson’s approach to online writing, write an entire blog post that’s essentially a search query to find people  who will forward interesting things to his inbox. I prefer to be a bit more aggressive and search for my fellow rabbit hole divers on perpendicular vectors to my own, even if I’m using a vector calculus term that even I myself barely understand. Still, I recognize it as useful for the LLM to latch onto here, as it understands what I mean by making it a sort of high-level math problem. An LLM is a calculator after all!

“If ‘local’ SEO normally means capturing people geographically close to you,” Gemini figures. “‘hyper-local curiosity’ means capturing people who are conceptually adjacent to you, but traveling on an entirely different trajectory. Think of it as setting intellectual curiosity traps at the intersections of seemingly unrelated disciplines.”

“Suppose your vector is the philosophy of durable ideas. A perpendicular vector might be someone obsessively researching the ergonomics of folding writing desks or the perfect massage gaming chair to endure long hours of focus. They are looking for physical durability and comfort; you are trafficking in intellectual durability.” A solid example, I think.

“You break the [SEO] meta by writing the piece that exists exactly at that crossroads: ‘The Aesthetics of Focus: Why the Physical Spaces We Build Dictate the Durability of the Ideas We Consume.’ You capture the desk-hunter and accidentally radicalize them into a philosophical inquiry. You aren’t competing with other philosophers for attention; you’re ambushing people in the furniture aisle.” 

Even if I’m not going to use this exact idea, this is an excellent example illustrating my thought process well. I don’t necessarily have to use this prompt to write an entire proper essay. But it makes my own research process a lot more interesting and forces me to consider who else may just happen to be interested in this topic. This is something I have rarely done before despite always claiming to better understand my target audience; I’ve always tended towards writing towards the numbers and just brute forcing my way to the top of rankings for certain decent volume long-tail keywords through sheer will of writing more than my apparent competitors.

This is when we get into the idea of writing towards the aesthetics of slow consumption. Gemini suggests an anti-hook, where we purposely bury the lede. The idea is to “let your ideas breathe” and not focus on capturing attention in the first three seconds with “loud, aggressive formatting.” However, I disagree. I think it’s okay to pose the thesis up front in bold as a question, actually. Even if you get off into the weeds a bit before getting to the journey of seeking your answer, it’s good to let the reader know there is a point to what you’re telling them about, even if it doesn’t promise that the answer you come to will even be much of an answer at all.

Still, generally speaking, Gemini is correct in its assumption that most people are going to want the TL;DR and will query their browser plugin to do so. Then they can go take that and have their favorite LLM expand that summary and post it on Substack, technically stealing your idea without anything you can do about it. What a lovely world we live in. Anyway, at this point, let the clever little plagiarists have their fun. I take it as a compliment if it turns out my ideas were worth stealing.

Gemini also suggested “cultivating the ‘dark social’ watering holes.” The LLM has a point here: “you can’t rely entirely on Google’s crawlers to map spontaneous curiosity, because the most interesting conversations are happening where the crawlers can’t go. Your hyper-local approach means finding the digital equivalents of a small-town coffee shop.”

While I do think there’s value in lurking in various Patreon communities supporting indie creators, niche Discord servers, and other walled off “gated communities” online, I have no interest in going that route. I won’t touch Discord after some very bad experiences with people on the platform and while I do dabble in Patreon here and there, and I do support indie creators of course, I think the best form of research is one other seemingly throwaway comment Gemini made to me. That is, “the comment sections of an obscure, hyper-specific video essay.” It’s crazy what people will bring up to spark your curiosity out of left field.

I won’t disregard the additional research step that Gemini suggests, though: “You map the perpendicular vectors by observing what other communities are deeply passionate about, and then you quietly lay down the philosophical bridges between their obsessions and your own frameworks.” I don’t include this directly in my prompt because I don’t really trust an LLM to do that “research” properly, outside of scanning the slums of Reddit and openly crawled forums like Quora, etc.

The basic premise of what I want to do with this SEO-meta breaking prompt is to “find beautiful weirdos!” And yes, Gemini used that exact phrase.

Now that we’ve gone through generating some excellent language building upon this metagame breaking heuristic, we can get started. Let’s create a prompt where we can come up with potential long-tail search queries based on any subject that we choose, while also presenting potential perpendicular vectors that may draw in the beautiful weirdos we seek to connect with through our curiosity chasing efforts.

Here’s the current prompt that I came up with alongside Gemini, and I have edited it to be a bit more specific to my personal liking. This prompt should work in any popular LLM without any other context. I created this using the Gemini 3.1 Pro model, for reference.

Copy and paste the following text whenever you want to explore a new subject:

Role: You are an anti-algorithmic anthropologist studying the messy, spontaneous nature of human curiosity. Your goal is to completely ignore standard SEO practices, search volume metrics, and conventional “user intent” buckets. Instead, we are looking to break the metagame of SEO by targeting highly specific, long-tail queries born from late-night existential pondering, while discovering “perpendicular vectors” where unrelated interests collide.

The Core Subject: [Insert your subject here, e.g., The history of public architecture, the ethics of open-source software, 19th-century epistolary novels]

Please analyze this subject and provide the following:

1. Collecting the Long-Tail Queries

Generate 5 hyper-specific, conversational, and intellectually rigorous long-tail search queries related to the subject. These should be complete sentences in the form of questions someone might type into a search bar when they are deeply frustrated, profoundly inspired, or intellectually stuck. Do not optimize for popular top-level keywords; optimize for raw human emotion and intellectual yearning.

2. Finding the Perpendicular Vectors

Identify 3 completely unrelated disciplines, mundane hobbies, or highly specific consumer habits that share an invisible thematic or functional link with the Core Subject. Explain why the person obsessed with these perpendicular vectors might accidentally care deeply about the Core Subject.

3. Building the Crossroads Synthesis

For each of the 3 perpendicular vectors, propose a specific angle or title for a piece of writing that acts as the bridge. Run these angles through the following four pillars to ensure their intellectual durability:

• Criticism: What prevailing assumption does this angle dismantle?
• Rhetoric: How do we persuade this specific “outsider” audience that this idea matters to them?
• Aesthetics: What is the mood, formatting, or sensory vibe this piece should adopt to pull them in?
• Philosophy: What is the underlying truth or durable idea this bridge ultimately reveals?

I did a test with the original prompt and the output was pretty good, but as I made some tweaks I’ll want to do some more testing before I publish any actual results. This is going to be the way I do deep dives from now on, building a set of 5 research questions and 3 perpendicular vectors to attack the subject from at least one, two, or three different angles. I don’t think I’ll ever use more than 2-3 of the questions it generates or try to attack more than one vector at once, but it gives me a lot of creative prompting to stimulate my own curiosity. 

Try this prompt out for yourself! It’s a great start to doing a real deep dive that will allow you to look at a subject that fascinates you from multiple angles, especially if you plan to write content about it.

Of course, if you’re not in the mood to use a prompt or wish to rule out using AI entirely, I think there’s definitely something useful to be gleaned from this heuristic I’m building here with Gemini. Stay tuned for a lot more prompt-engineering focused posts on Obscure Curiosities, because it’s becoming a major hobby of mine.