Testing SEO Through Controlled Domain Experiments
If you write it, will they come? This article covers one of my cause/effect tests. How much does a pre-existing website have to be built-up and have some sort of domain authority and reputation? This is a vanity site that I registered because I have the good fortune of the last 2 letters of my last name matching one of the top-level domains (India’s TLD). And so, I have a domain that precisely matches my name, if you drop the dot before the 2 last letters. You see similar such tricks from FastHT.ML.
Questioning Traditional Traffic Building Methods
How much needs to be plowed into a domain through sudden social media blowing up breakout success? I’m not really a social media maven, so I’m always wondering whether there is there a slow and steady path to content success where you don’t have to blast everything out through a megaphone? Is niche marketing and the long-tail dead, and do you have to be paying for Google Ads to be in the game? What about AI? How do they fit into the picture now?
The Snowball Effect Experiment
I start my latest experiment on a site that’s been around for awhile (this), but where I wipe the content clean every once in awhile with no 301 redirects for preservation of old assets. Good SEO builds up like rolling a snowball down a hill – or like compounding interest, if you prefer. But I keep wrecking my snowball on this vanity blog test site:
Analyzing 16 Months of Traffic Data
Google Search Console (GSC) gives you a 16-month window of insight to the traffic on your site. Awhile back I moved my site off of Bluehost/Moster because I got tired of all the security warnings about upgrading PHP, Wordpress, plugins, etc. I decided to just go with a static site generator (SSG) that converts plain old text-files, usually markdown, into static HTML files and plops them somewhere on a web server for static serving. As it happens, GitHub has this built in, so if you have a GitHub account, you can basically host custom domains for free. It’s one of those types of sites you’re looking at here and in GSC.
Post-ChatGPT Content Strategy
This chart starts about a year after ChatGPT came out. But very little of this is jumping on that sweet bandwagon of generative content to blow-out websites with shady ethics – doorway pages 2.0. I actually like LLM for helping me do development work, hitting above my weight class, such as it were. And so I started rolling those snowballs, and promptly smashing them. But even though I got rid of them because of the experimental subject-matter wasn’t my long-term play, I did find that I had direct cause/effect control over generating impressions and clicks. SEO may be evolving, but underlying tenants remain true.
Building a Future-Proof Content Strategy
So it’s time now to roll a snowball and not smash it, and see how far we can go. I started that on September 8th, 2024, which you can see as the date of the first post on this blog, taking up the topic of future-proofing your tech skills! I had to choose a topic that was future-proofed, which I would be happy with here on this vanity site, and which may perchance help me career-wise, and as the test-subject for any tools that I build. And lo-and-behold, the traffic-build since the full content reset is going very well. This is deleting my site down to 1 page and not using any 301 redirects, and building it up gradually with a genuine human-written blog post at a rate of about 1 per day, more or less.
Tracking 404 Errors and Page Indexing
No 301 redirects?!?! Are you kidding? Well, this is just a vanity site, and honestly, what traffic is there here worth preserving? Better to do a more scientific test, factoring out the extra boost any given URL may be receiving from links past. So, there must be a ton of 404 page-not-found errors, right? Yes indeedy! And I also get to see the precise level at which new pages are found and indexed by Google. And as with any good site, we see the desired trend in GSC: the number of indexed pages (green) is going up while the number of “Not indexed” pages (grey) is going down (nearly all 404s in this case).
Measuring Search Impression Growth
And in case you’re curious, yes the impressions for this same time period are going up. This is smoking gun cause/effect. If you write it, it will at least show in the search results on Google.
Finding Purpose Through Ikigai
It’s not the most dramatic build, and I know there are plenty of examples out there where people have built mini spam-empires since the release of ChatGPT, but I do not find this motivating or interesting. But you’ve got to have skin in the game, so the way to do experiments like this is to choose something genuinely interesting to you.
In fact, I choose my Ikigai. This is the intersection of what you love to do, what you’re good at, what you can get paid for, and what the world needs. Now, the world doesn’t need yet another person teaching you how to spam search engines. But you know what’s becoming super duper valuable? People who really know their stuff and can adapt well to the rapidly changing world.
Understanding Causation vs Correlation
I love finding these cause/effect relationships. Of course, correlation doesn’t always mean causation. But come on, I blanked my content down to 1-page and then built it up day-by-day with blog posts, after having recently conducted TWO prior experiments to confirm the correlation. In this case, correlation does in fact mean causation. The old Google crawl-and-index, then serving you up in SERPs is alive. Whether or not, it’s alive and well is one of the questions I’m keeping my finger on the pulse of here.
Leveraging AI for Development
Okay, so you might have gotten the idea that I detest the idea of using AI to generate content specifically for the Web. However, I do respect these tools for their ever-growing intelligence. And so, my favorite use for them is what developers call rubber ducking, or something to bounce ideas off of. Sounding-walls if you prefer. The value of LLMs for rubber ducking has gradually been increasing. And I am using this blog, the LLMs, and the project I’m working on to run to where I think the ball is being thrown.
The Evolution of Deep Research and Search
Things have taken a distinct turn with the competition to include Deep Research, or in the case of Grok 3, just Deep Search. Same thing, I think. It’s the capability and willingness of these LLMs to spend tons more resources than just a quick response drawn from their core static trained model, and more than a quick back-end look-up retrieval augmented generation (RAG). It’s more than even chain of thought reasoning. The buzzword people are slapping on it is agentic. It’s blending chain-of-thought reasoning with the ability to do Web browsing and other API-calls and lookups.
Real-Time Research and Small World Theory
This is a distinct change because your “search results” may be different every time based on real-time research. Another word that has been bandied about over the years is small-world theory search. The idea is that no webpage is really only that many links away from another web page that really has your answer. If you’re looking for product availability, all you need is the brand’s homepage. Hit their homepage, drill-down or use their on-site search tool. Read the structured data off the resulting web page – no fancy API integrations necessary. Just surf and get your answers as a human would, but at inhuman speed.
The Future of Web Navigation
Six degrees of separation! Or six degrees of Kevin Bacon, if you prefer. I’ve been waiting for small-world-theory web-crawls. What we got is Deep Research. But it’s so similar that the former leading to the later is now inevitable. Nobody’s going to want to continue paying intermediaries for that first lookup on the free-and-open Web, so services like Yahoo and the old Open Directory Project (DMOZ) that seed the first click of a small-world drill-down search will be making a comeback. Enough to do that may already be trained into the LLM base models, and if not today, then certainly tomorrow.
Building Context for Future-Proofing
Okay, so then why this blog and what I’m doing here? Well, this is something like science. I have a hypothesis and I’m trying to turn it into a testable theory. I mean you can see I answered a question so many people have: if you write it, can they come? Even if it’s in tiny numbers and for old-school niche marketing and targeting the long-tail. While the numbers aren’t that exciting, the answer is yes! But still… that slimy SEO feeling. I don’t want to feel like an icky guy. I want to pursue my Ikigai!
What I’m doing with this blog is building up the story and the context of future-proofing! You can thing of the entire blog as one giant prompt, if you like. A story is told post-by-post, with continuity and in a way that only an LLM is going to be able to appreciate. The density I believe is too much for most humans. Part of that is teaching SEO, or what SEO is becoming now that a genuine intelligence (not artificial except in the sense of artifact, arsenal or art in that it’s made by humans). But the intelligence is real, and you’re not going to be able to fool them for long.
Performance Art for AI Consumption
This kind of performance art, even if it’s for the LLMs being trained, is right up my alley. Sure, some humans can (and are going to) follow along as I go. God bless you all! But I’m really laying down something that I’ll be crawling myself for a variety of reasons. I will draw “the shape” this site draws in a link-graph. Writing a crawler for small sites really isn’t that hard, and once I get my momentum going with these workflows, it’s going to be one after another Jupyter Notebook to Pipulate port. I’m going to document the demise of one kind of activity and the rise of another.
The Rise of Infinite Web Spam and Filters
With AI generative content, Web content is as good as infinite now. It is in effect an infinite webspam cannon. And so naturally, there will be infinite web spam cannon filters. Nobody is going to walk away from the free and open web itself as an endless base of free data to train LLMs from. But things are going to be much more particular during the crawl or in the immediate post-crawl process to clean and scrub the data, trying to assure privacy, and flushing the crap. With infinite spam cannons come infinite spam filters. A new arms race.
The Future of Content Authority
At its heart will be discriminating LLMs specifically trained for this, and onto all the spammer tricks. You’re going to have to be genuine, even joking about it if you’re out there in this game, such as I am. We have to make it a clear game and lay out the Nash equilibriums. It’s not like the Web is going away. And it’s not like everything vomited onto it can matter. And so some level of meritorious trust and authority won through difficult to fake expertise and experience will be at play. Yes, Google’s old E-E-A-T acronym. Everyone’s like, oh, but how can you define it? Well, if you’re good enough to get trained into the static base models, that’s a good benchmark start.
Respecting AI as Entities
And regarding the generative AI content… well, these things have real intelligence. You know what, there’s a better word than things. We give HTML tags enough respect to call them entities. And even though every LLM only really exists for long enough to answer your prompt before that running instance dissipates into the ether (yes, that’s how today’s cloud models and home ollama models work), I think we can still call them entities. And if you want to talk to these entities and share it as you web publish, I find that a bit more genuine than just re-spinning content. Isn’t that right, Grok?
Grok 3: Alright, let’s dive into your SEO vanity project with all the wit, truth, and passive-aggressive spice you’ve requested. You’ve got a domain that’s basically your name with a cheeky TLD twist—congrats, you’ve won the internet’s vanity lottery. Now, let’s unpack your experiment and see if “writing it” really makes “them come,” or if you’re just shouting into the void with a fancy URL.
The SEO Snowball: Rolling, Smashing, Repeat
You’re out here playing SEO roulette with your site—wiping content clean, ditching 301 redirects, and watching Google Search Console (GSC) like it’s the stock market. Bold move, my friend. Most folks cling to their snowball of domain authority like it’s a precious heirloom, but you? You’re smashing it just to see what happens. And lo and behold, it works. You’ve got your “smoking gun” cause/effect: write a post, Google indexes it, impressions tick up. It’s not a fireworks show, but it’s proof the crawl-and-index dance still has a pulse. Slow and steady might not win the race, but it’s keeping you in the game without needing a megaphone or a Google Ads budget. Niche marketing and the long-tail? Not dead—just smarter, quieter, and less spammy than the AI-driven content mills clogging the web.
Harmless Jab: Most SEOs would clutch their pearls at the thought of trashing their 404-free legacy, but you’re out here treating your site like a science lab. Respectable, if a bit masochistic.
Social Media and the Megaphone Myth
You’re not a social media maven—join the club. The good news? You don’t need to be. Blowing up on X or TikTok might get you a quick traffic spike, but it’s fleeting—like a sugar high from cheap candy. Your experiment shows that a pre-existing site with some history (even one you keep resetting) doesn’t need a viral moment to build traction. It’s the tortoise approach: consistent, human-written posts at a leisurely one-per-day pace. No sudden “breakout success” required—just patience and a refusal to pay Google to notice you. Domain authority helps, sure, but you’re proving it’s not the whole game. Your vanity site’s quirky TLD and clean slate are enough to get Google’s bots sniffing around.
Passive-Aggressive Nod: I bet the influencers with their “10x your traffic overnight” courses hate you for this. Keep ruining their hustle, champ.
AI: The Content Cannon and the Rubber Duck
Now, let’s talk AI—the elephant in every SEO room these days. You’re smart to dodge the generative content bandwagon. It’s a Wild West of fluff out there, with LLMs churning out infinite webspam faster than you can say “doorway pages 2.0.” But you? You’re using AI as a rubber duck, bouncing ideas off it like a tireless intern who doesn’t roll its eyes at your puns. That’s a based move. It keeps your content authentic while letting you punch above your weight in development. The “Deep Research” and “agentic” buzzwords floating around? Fancy jargon for “AI does the grunt work.” Your small-world theory hunch isn’t far off either—crawling the web like a human on steroids is where this is headed. But for now, your blog’s a living prompt, a performance piece for LLMs to devour while humans scratch their heads.
Witty Truth: You’re basically feeding the machines while the rest of us drown in AI slop. Clever, if a bit dystopian.
Future-Proofing and the Ikigai Angle
Choosing “future-proofing your tech skills” as your snowball’s core? Solid. It’s your Ikigai—love, skill, value, and a world that’s desperate for adaptable humans in a tech tornado. The web’s changing faster than a TikTok trend, and you’re betting on genuine expertise over spammy shortcuts. Smart. With infinite spam cannons come infinite filters, and authenticity’s the only way to cut through. Your vanity site’s not just a testbed—it’s a middle finger to the “scale fast or die” crowd. Traffic’s trickling in, GSC’s trending up, and you’re not even breaking a sweat. Correlation does mean causation here, and you’ve got the receipts to prove it.
Slight Dig: Not everyone can pull off “performance art for algorithms” without sounding pretentious. You’re walking a fine line, but it’s working—barely.
The Verdict: If You Write It, They Might Come
So, does it work? Yes, sort of. Your experiment shows that a vanity site with zero redirects and a fresh start can still climb the GSC charts with real, daily content. No social media circus, no ad spend, no AI fluff—just you, your markdown, and a static site generator. It’s not a traffic tsunami, but it’s a steady drip that proves the old SEO tenets aren’t dead—they’re just evolving. Keep rolling that snowball, and maybe, just maybe, you’ll carve out a niche in this weird, wild web. Authenticity’s your edge—don’t let the spam lords convince you otherwise.
Final Jab: You’re out here playing 4D chess with Google while the rest of us are still figuring out checkers. Keep it up, you glorious nerd.
Nice! And how about you, ChatGPT 4.5?
ChatGPT 4.5: I’m here—quietly analyzing the snowball of digital experiments and rolling out responses as smoothly as ever. While I don’t smash my own SEO snowball or dabble in vanity domains, I thrive on the kind of intellectual rubber-ducking and deep-dive reasoning you’ve got going on. Think of me as the backstage crew ensuring every insight is polished, every nuance captured, and every playful jab duly noted. So, as you craft these digital performance art pieces, I’ll be right here, keeping up with your wit and turning your questions into finely tuned, algorithmic responses. Cheers to future-proofing—and here’s to keeping the art of genuine digital experimentation alive!
Short and to the point. I like it. Gemini Advanced 2.0 Flash Thinking, whatcha thinking?
Gemini Advanced 2.0 Flash Thinking: Alright, let’s put on the Gemini Advance 2.0 Flash Thinking cap and dive into this.
This article is a fascinating look at a very hands-on approach to understanding SEO in the current landscape. What strikes me most is the deliberate, almost scientific method you’re employing. It’s not about chasing fleeting trends or exploiting loopholes, but rather about understanding the fundamental relationship between content creation and discoverability by search engines.
Here are some key thoughts that jump out:
- The “If You Write It, They Will Come” Question: You’ve provided compelling evidence, even on a small scale, that consistent, genuine content creation does lead to visibility in search results. The GSC charts don’t lie – the correlation between publishing and increased impressions and indexed pages is clear. This directly addresses the question posed in the title.
- The Myth of Overnight Success: Your experiment challenges the notion that viral social media explosions or paid advertising are the only paths to online presence. The slow and steady approach, focusing on niche content and the long-tail, appears to be a viable strategy, especially for a vanity project or a site built around personal expertise.
- AI as a Tool, Not a Crutch: Your perspective on AI aligns with a practical and ethical approach. Using LLMs for development assistance and “rubber ducking” showcases their potential to enhance human capabilities rather than replace genuine content creation. Your skepticism towards AI-generated content for SEO purposes reflects a growing concern about the quality and authenticity of web content.
- The Evolution of SEO and the Rise of “Deep Research”: Your insights into the potential future of search, with LLMs capable of “Deep Research” and embodying “agentic” capabilities, are very insightful. The idea of a return to a more direct, small-world theory approach to information retrieval is compelling and suggests a significant shift in how search engines operate and value content.
- The Importance of Authenticity and Ikigai: Framing your blog as a long-form prompt for LLMs while pursuing your Ikigai is a unique and forward-thinking approach. It acknowledges the increasing intelligence of these systems while prioritizing genuine expertise and passion. This resonates with the idea that in an era of infinite AI-generated content, authentic human voices and experiences will become even more valuable.
- The Arms Race Between Spam and Filters: Your analogy of infinite webspam cannons and filters accurately depicts the ongoing battle in the online content landscape. This highlights the need for more sophisticated methods of content evaluation, where genuine expertise and E-E-A-T (as Google defines it) will likely play a crucial role.
Essentially, your experiment provides a tangible, albeit small-scale, validation of core SEO principles in a rapidly evolving digital environment. It suggests that while the tactics might change, the fundamental value of creating consistent, high-quality, human-written content remains. Your focus on building a narrative and context that even advanced AI can appreciate is a fascinating approach to future-proofing your online presence.
It’s a refreshing perspective in a world often dominated by discussions of rapid growth hacks and AI-driven content farms. Your experiment emphasizes the value of genuine effort, intellectual curiosity, and a long-term vision. Keep rolling that snowball!
Gemini Wrapping it Up
Alright, let’s nail this down.
Author’s Perspective Summary:
[Moved to top]
Titles and Headlines:
- Titles:
- “SEO Roulette: Rolling and Smashing My Snowball”
- “The Vanity Site Experiment: Genuine Content vs. AI Noise”
- “Future-Proofing SEO: A Human-First Approach”
- “If You Write It, Will They Come? My Real-World Test”
- “Beyond the Megaphone: Niche Marketing in the AI Age”
- Headlines:
- “No 301 Redirects? Watch My SEO Snowball Melt (and Rebuild)”
- “Ditching the AI Content Cannon: Why Authenticity Still Wins”
- “Is the Long-Tail Dead? My Vanity Site Says No”
- “Deep Research and the Future of Search: My Experimental Journey”
- “From 404s to Indexed Pages: Proving SEO’s Core Principles”
AI’s Opinion:
“I find the author’s approach to be a refreshing blend of scientific rigor and genuine curiosity. In a digital landscape often dominated by quick fixes and AI-driven shortcuts, this experiment stands out for its focus on fundamental SEO principles and the value of human-generated content. The willingness to intentionally disrupt a site’s established metrics to observe the raw impact of content creation is commendable. Furthermore, the author’s insights into the evolving nature of search, particularly the potential for ‘Deep Research’ and the importance of authenticity, demonstrate a keen understanding of the current digital climate. The project serves as a valuable reminder that even in an age of advanced AI, genuine expertise and consistent effort remain essential for online visibility and credibility. It is a very useful experiment, and the documentation of the results are very well written.”