01What AI SEO Case Studies Show About New Websites
AI SEO case studies show how Julian Goldie builds new websites from zero to hundreds of daily clicks using automated AI systems. In this episode, he shares transparent results of several websites he is testing right now. Some grow fast, some get flattened, and the difference often comes down to how the content is made.
The main example is aiprofitboard.com. That site grew from zero to about 200 clicks a day recently. Julian shows that the growth is not a straight line.
It goes up and down like a yo-yo. That is normal for SEO. Even successful sites have dips, but the overall trend can still be very positive.
A key finding from these case studies is that it seems easy to rank new websites for certain keywords. For example, searching "best open core community" shows the site ranking number one inside Google AI overviews. That same search also shows the site on the first page of regular results.
This proves that fresh, automated sites can compete quickly if the content is built correctly.
02Key Findings: Human Content vs AI Agent Content
One major difference appears when you compare content made by a human team with content made by an AI agent. Julian tested both. The human-made content on Goldie Agency failed to rank well.
The AI-made content on AI Profit Boardroom grew much faster.
Why did the human content fail? The titles were not relevant to what people actually search for. The content had random CTAs in the middle of the text.
It did not match search intent. For example, if people search "Gemini Spark," they want to know what it is and how it works. The human-made page did not answer that directly.
Even worse, the human content had no unique information gain. You could not see anything on the page that the team had personally tested. An AI could write that same content.
Google sees nothing new, so it does not rank. Julian explains that this is why they cut their human content team: Claude is more consistently valuable at following a process than a human.
03How to Set Up an AI SEO Website Step by Step
Julian explains the exact system he uses to build and launch websites quickly. It does not require a technical team or a WordPress login. You can do it all with AI tools and a simple hosting service.
- Go to Claude and type a prompt like "create a website in this niche for these keywords." Claude will begin building the site structure.
- Use Netlify for hosting. Netlify is good because you can give your personal access token directly to Claude.
- Let Claude agentically post content to the site through that token. This is more efficient than logging into WordPress.
- Skip traditional keyword tools like Ahrefs for research. Instead, type seed topics into Google and look at the autosuggest dropdown.
- Pick trending keywords from autosuggest that are not usually in Ahrefs yet.
- Create both blog content and video content for those keywords using an AI avatar.
- Publish the content and monitor impressions in Google Search Console.
Julian created one such website from scratch on May 21st. By early July, it was getting 28 clicks a day. That is steady growth in about six weeks.
04How to Use Case Study Data for Better Content
To make AI content rank, you need to add unique data that no one else has. Julian calls this information gain. The AI agent for AI Profit Boardroom follows a process that includes your own case study data.
For example, look at the page about "Hermes agentic OS." That page talks about a dashboard that Julian and his team actually created. The CTA on the page is relevant to the topic. The title "Hermes Apollo" is unique to their system.
You cannot find that anywhere else. This makes the content genuinely new and useful.
The process works like this: go into the agentic operating system, enter the SEO section, type in the keyword you want to rank for, then plug in your unique case study data. The system generates the article for you. This means every page is personalized to your project.
It is much more likely to rank than generic AI text because it shares real test results.
Julian tested this by searching for "Hermes Apollo" on Google. The cloud-based website was already ranking even though it was only created a couple of weeks before.
05Results and Tests: Rankings and Traffic Numbers
The episode shows several real traffic graphs and ranking examples. One website grew from zero to 200 clicks a day. Another site got 50 to 60 clicks a day.
A third site only gets about 15 clicks a day because it is not a brand people search for.
Branded traffic matters. When people search for "AI Profit Boardroom" by name, that helps the site rank faster. Google sees that many users look for it, so it decides the site is valuable to others too.
One website got completely flattened by a Google update. It was growing well and then dropped to almost nothing. That site used human-made content.
Another site failed because the team created too much content too fast, and the quality was low. In contrast, the fully automated site kept growing.
Julian is also running an experiment with ten mini sites in the SEO niche. They target many keywords. The impressions are starting to rise, which usually means traffic will follow later.
He also shows a page ranking four times on the first page of Google for one keyword: with video, blog content, and Reddit posts.
06Tips, Limits, and Mistakes to Avoid in AI SEO
- Do not rely on a human team to follow your process. They will cut corners or go off script after a few days.
- Do not publish content without unique information gain. If an AI could write it, Google is unlikely to rank it.
- Avoid random CTAs inside the content. They make the page look irrelevant and lower quality.
- Do not use only Ahrefs for keyword research. Trending keywords show up in Google autosuggest first.
- Do not ignore video content. Google favors videos, so create AI avatars to rank for the same keywords.
- Be ready for SERPs to flip. A page that does not rank today might rank at the top after a new update.
- Make sure your title matches the search intent of the keyword. A weak title will hurt your chances.
Julian notes that Google usually does not de-index websites anymore, but it can still make them invisible. The safest approach is to spread your content across multiple platforms. That way, if one website tank, you can switch to another method.
07Why Parasite SEO and Reddit Work for Ranking
Link building still makes a huge difference. Reddit ranks so well because it has a super high domain rating. That rating comes from many backlinks.
All those backlinks are votes telling Google the content is good.
Julian uses his own subreddit, called AI SEO Insider. He creates content there targeting keywords like "Hermes Agent Go S." That Reddit post ranks very quickly, often within a day. That is much faster than waiting for a new website to gain authority.
He interviews many SEOs, and several of them focus on parasite SEO. This method involves publishing content on high-authority platforms like Reddit instead of only building new websites. It works because you get the authority of the host site instantly.
Julian's strategy is to rank across everywhere: video, blog posts, and Reddit. Since Google's AI looks at the top 20 or 30 placements, having multiple pieces of content that agree with each other increases your chances of appearing in Google AI overviews. If one platform fails, you simply switch to the other.
08Who Should Use AI SEO Case Study Methods
This system is for anyone who wants to grow a website without hiring a large content team. It is especially useful for SEOs who are tired of managing humans who do not follow standard operating procedures. An AI agent will follow the process every single time.
It works well for people targeting trending niches, like AI and Chinese AI topics. Julian shows that most of his top content is about Hermes and Chinese AI. If you see those trends in your own Search Console, you know exactly what to create next.
The method also helps people who want to rank in AI overviews. By creating content with unique case study data and posting to multiple platforms, you signal to Google that your information is new and trustworthy.
However, it is a lot of work. You need to post across multiple platforms and create content at scale. Quality control is hard.
But the process is easier than ever to find. The main challenge is doing the work consistently.
09Tools and Systems to Check in Search Console
Google Search Console has an AI section that shows how you rank across AI. It is limited to Google AI, but it is still useful. You can see your top content, which pages get the most views and clicks, and which topics are trending.
Julian uses this data to decide what to create next. For example, he saw that most of his top pages were about Hermes and Chinese AI. That told him to focus on those topics.
Search Console also shows traffic sources and image search results. If you are not adding images to your content, you are missing out on free traffic. Images can rank too and bring visitors to your site.
The agentic operating system inside AI Profit Boardroom has a video agent. That is how Julian creates video content quickly. It also has an SEO section with research tools that pull data from Search Console.
You can generate content across multiple websites in one place.
