# Make 3k SEO Pages from 6 Templates with Claude Code and MCP

## Metadata

- **Published:** 7/11/2025
- **Duration:** 20 minutes
- **YouTube URL:** https://youtube.com/watch?v=iR6o19zvZ7o
- **Channel:** nerding.io

## Description

Join the Community: https://nas.io/vibe-coding-retreat

Achieve massive organic traffic with programmatic SEO and #mcp ! In this tutorial, we'll show you how to generate 3,000+ unique landing pages from just six templates using Claude Code and structured data—ideal for directories, locales, or product catalogs.

🛠 What You’ll Learn
- What programmatic SEO is and why it's essential for scaling long-tail keywords 
- How to design six adaptable templates for dynamic page generation
- Using Claude Code to fill those templates at scale
- Integrating data sources—API, MCP, or database—for rich, contextual content
- Best practices to avoid thin content and maintain SEO quality 
- Tips to monitor and optimize generated pages over time

💡 Why It Works
Programmatic SEO lets you automate thousands of targeted landing pages, capturing niche, high-intent search traffic without manual effort . With Claude Code, you bring AI-driven content generation into the process, boosting efficiency and content quality.

## Key Highlights

### 1. Autonomous Keyword Expansion with Claude Code

The video demonstrates using Claude Code to autonomously expand keywords and enrich data for programmatic SEO, building upon a pre-existing system. The goal is to generate dynamic URLs that connect with the Google SEO algorithm.

### 2. Leveraging Custom MCPS for SEO

The presenter uses custom MCPs (presumably Model Configuration Presets) within Claude, including one inspired by the "boring marketer" to guide the AI's SEO strategy and coding process. These MCPs grant access to various tools and API keys.

### 3. Overcoming Challenges with API Keys and Tool Integration

The process involves debugging API key issues, specifically with perplexity, and adjusting the AI's workflow to effectively utilize tools like Firecrawl for data scraping and content generation, showing a real-world development process.

### 4. From 6 Templates to 2,000 Pages

The key achievement is expanding from a small set of templates (5 or 6) to generating approximately 2,000 SEO pages through a bottom-up longtail funnel approach, significantly increasing the potential search visibility.

### 5. Iterative Refinement of AI-Generated Content

The process isn't fully automated; it requires manual review and fine-tuning of the AI-generated keywords, content, and structure to ensure alignment with the overall SEO strategy and desired outcomes.


## Summary

## Video Summary: Make 3k SEO Pages from 6 Templates with Claude Code and MCP

**1. Executive Summary:**
This video demonstrates how to leverage Claude Code, MCPs (Model Configuration Presets), and programmatic SEO techniques to generate thousands of unique landing pages from a small set of templates. The process involves automating keyword expansion, data enrichment, and content creation to capture niche, high-intent search traffic with minimal manual effort.

**2. Main Topics Covered:**

*   **Programmatic SEO Fundamentals:** Explanation of the concept and its importance for scaling long-tail keywords.
*   **Template Design:** The role of adaptable templates in dynamic page generation.
*   **Claude Code Integration:** Utilizing Claude Code to autonomously generate content at scale.
*   **MCPs (Model Configuration Presets):** Customizing AI behavior for SEO strategy through MCPs.
*   **Data Source Integration:** Connecting to APIs (Perplexity, Firecrawl, Data SEO) for richer content.
*   **Keyword Expansion:** Demonstrates using Claude Code to autonomously expand keywords and enrich data for programmatic SEO
*   **Debugging and Problem Solving:** Addressing API key issues and adjusting AI workflows.
*   **Iterative Refinement:** The importance of manual review and fine-tuning of AI-generated content.
*   **Bottom-Up Longtail Funnel:** Generating pages from the "bottom" by identifying long tail keywords

**3. Key Takeaways:**

*   Programmatic SEO can significantly expand organic reach by automating the creation of targeted landing pages.
*   Claude Code can automate keyword expansion, content generation, and data enrichment for SEO.
*   MCPs (Model Configuration Presets) allow for customized AI behavior tailored to specific SEO strategies, enhancing efficiency and control.
*   API integration with tools like Firecrawl and Perplexity can provide valuable data for content creation.
*   The process requires iterative refinement and manual review to ensure content quality and alignment with SEO goals.
*   A small number of templates can be scaled to create thousands of pages, allowing for massive visibility.
*   Generating from the bottom-up (long tail keywords) can create more pages than traditional methods.

**4. Notable Quotes or Examples:**

*   "What that means is you have these URLs and they are like dynamically getting generated and we do this and the outcome is this and what we do is we have a call to action and then basically some kind of sales pitchy landing pages that will connect on the Google SEO algorithm." - Explanation of programmatic SEO implementation.
*   "Using different types of MCPS, we could actually do a keyword analysis and actually expand the data to a much higher range." - Highlighting the power of MCPs in scaling SEO efforts.
*   "From five or six base pages, we then created 2,000 pages. That's just that's that's wild." - Demonstrating the scalable outcome from the process.
*   Leveraging custom MCPs (presumably Model Configuration Presets) within Claude, including one inspired by the "boring marketer" to guide the AI's SEO strategy and coding process.

**5. Target Audience:**

*   SEO professionals looking to automate and scale their content creation efforts.
*   Marketing professionals interested in leveraging AI for programmatic SEO.
*   Developers seeking to integrate AI-powered tools into their SEO workflows.
*   Entrepreneurs and business owners aiming to improve their online visibility and organic traffic.
*   Anyone interested in the intersection of AI, SEO, and content creation.


## Full Transcript

Hey everyone, welcome to Ning.io. I'm JD and today what we're going to go through is we're going to take the programmatic SEO system that we previously built and we're going to use claude code to actually extend the keywords and build out more of the data in an autonomous way. With that, let's go ahead and get started. All right, so I'm continuing my journey in pro programmatic SEO and I am going to I'm focusing on the vibe coding retreat. So again, this is a uh local cohort that I'm putting together where basically I'll be in different cities and want to connect with other builders where we can go and build something. The idea is that you have your concept, I'll help you code the last 10% and get everything over the finish line. So all of this has been coded in uh B 0 and cursor. And in the previous uh videos I've kind of gone through exactly what I'm trying to do where I have this concept of multiple different cities. You can register and vote. And so I want to take this and put it into a dynamic program programmatic SEO system. What that means is you have these URLs and they are like dynamically getting generated and we do this and the outcome is this and what we do is we have a call to action and then basically some kind of sales pitchy landing pages that will uh connect on the um Google SEO algorithm. So again this was all all of this was done in cursor and uh v 0. So now what we're going to do is we're actually going to use claude code and we're going to take inside of cursor we're going to take claude code and we're going to kind of give it instructions. I'm going to make what's called use MCPS and we're going to kind of build an SEO system that's actually going to go out and like pull information from uh keywords and actually try and expand this a bit better. Real quick everyone, if you haven't already, please remember to like and subscribe. it helps more than you know. If you're looking for on-site training, please click the vibe coding retreat down below. And with that, let's get back to it. So, in order to do that, I'm going to go into cursor and we can look at the PRD that I put together. So basically what I did and from the last video I had this strategy and from the strategy I was able to take information that had kind of done and start putting things into uh a programmatic SEO instructions. So I'm taking this programmatic SEO instructions and the combination of this SEO strategy and I am going to start doing uh the uh PRD. And so you can kind of see that in my um my prompt. I basically took this I started with a starting point of whatever prompt I want. I literally just told it to spruce itself up and use the uh instructions as an example from that. I'm now going to take this and run it through cloud. So, I haven't really done this before um and we're going to see what comes out of it. Key things that you should note is that I've already defined what MCPS I want to use. I know what structure I have a reference to these markdown files and where they're stored, right? So, they're at the root. it knows where my programmatic SEO library is and the uh data that I actually want to send here. So, we're kind of looking at this kind of information and we're going to start uh running this through the cloud code. So, what I'm going to do uh is not sure why it did that. It's taking a second. Okay. So now we have um our clock code and the first thing I want to point out is I've already gone up and and set these up. So you have this ability to set different MCPS. So all of them are connected. Uh, I've created the boring marketing which I'm gonna say actually in here use boring marketer as a check red. Okay. So, what this is is it's actually a a custom MCP that I made where you in one of one of my last videos where you can actually take your favorite YouTuber and uh use their strategy in your coding process essentially. Um, so I'm I'm trying I'm learning a lot from boring marketing uh and uh yeah want to continue using that. So the other thing is contact 7. So these two are actually coming from uh my MCP that's in claude. Then this is actually coming from the MCPS that I have set up inside of Claude. I think I might have said that backwards. This is cursor. This is claude. And so but it has access to all of these. And each one of these has different tools. So if we look at like uh this is telling us what tools we have. So there's a search tool. We can escape to go back. What uh I don't like firecrawl and perplexity and data SEO all have API keys. I won't show those. But the idea is that they they all have different tools. Let's see. And so this has like 49 different tools that it can leverage. So I'm hoping that it'll pull from this in order to get as much of the keywords and everything else that it can in order to run this. Okay. So now what I'm going to say is uh we'll just say do execute the SEO instructions for this P ID and let's come over here and get the relative path and we'll go ahead and see what happens next. It's going to ask me throughout tool name custom name string should have at most 64 characters. All right. So let I'm going to go debug this. Do a keyword analysis. We'll go back and we'll grab our code. We'll add some context of our cities again use every to possible conduct. We'll see. All right. So, it's going and looking and it's getting a bunch of these. It did not actually make them. It did go out and find Asheville. Found some other keywords. It is actually putting them in there. Uh, so we could probably change this to JS trying to get like more information out of here. Uh, but once I get Great. Okay. So, let's change this back to a markdown and we'll say apply here. And then what we're going to do is we're going to go back into here and we're going to say use and and then we'll say our path here and see if added words. Great. Okay. So now we have like a lot of different keywords, a lot of different information. Um, we're able to pull this back, right? So now that we have our keywords, let's go ahead and open claude and we'll do uh we'll try this again. Let's go ahead and accept this file. I think we need to accept this one, too. Nope, just bugs. Okay, cool. All right. So, let's make sure that we have the MCP turned off for data SEO. We do not. So, what we're going to do here is if you go quad MCP remove then that will actually remove it from there. And we can we're going to have to set this up again. So claw and now we can actually run this. All right. So we can just press up and then we can try this again and let's see if we have better luck this time. All right. So it's actually reading the uh extended. It read the keywords. Cool. So we actually have the pro. It's pulling in the pro SEO instructions. Now, it has a it's putting together a comprehensive to-do list. So, I'm just going to go ahead and let this run for a minute. Okay, cool. So, as you can see, it's got to-dos. So, it's going to do comprehensive research on the cities. Research all tech ecosystem for each city. Then, it's going to research city specific advantages, neighborhoods, cost of living. I'm not really sure why it's doing that. Um, but we're going to let it run. So it is doing a lot of analysis. Um once it starts to settle I'll unpause. All right. So this is the first action that it came back to me with. It's actually looking for perplexity to do its research. Uh as you can see it also did research over here. So we're going to say uh that yes and not ask again. So, it's just going to run the perplexity search from now on. And I'll pause this while it's freaking out. All right. After some turmoil, I keep running into an issue where it's getting stuck on perplexity. So, what I'm going to try and do is tell it to adjust itself. Uh, cool. Now, we're getting an unauthorized. Looks like I might need to update uh my API keys. All right. So, finally got this working a little bit. Now, it's going out and actually doing fire crawl. I had to do uh the competitive analysis was getting hung up. So, now we're actually doing fire crawl, which will actually like scrape uh different areas and pull back information and then actually build Cool. So, I'm going to let this run and uh see now that we actually are getting queries back information about There we go. Cool. So, now it's going through and we're creating the uh uh JSON. So, we're just going to go and say yes. No, ask anymore. when we're building out the uh next data sets. So, I don't think it'll ask, but I'm just going to let this run. All right. So, now it's completed and we've got all the to-dos. So, the project summary now is that it's focusing on specifically the fintech programmatic. It did a compre comprehensive on that. It fixed the enhanced SEO infrastructure. It did all of this research. And actually, it even did a SEO performance report, which I have no idea where that is, but I definitely want to see it. Uh, and see if we can get that loaded somewhere. Cool. Interesting. So put in the search volume, put in the different pieces, uh did evaluation on different boot camps in the area and a lot of different how to longtail how-to guides. Uh still keeping the same structure. Cool. So I bit close but uh since it completed you know where we can go from here is basically taking that and trying to uh load in all of this information. Again looking at all these different topics. I'm going to go through and try and make sure manually and just kind of look at the different types of projects, different types of uh tools that it's trying to specify and see if it aligns with what I'm looking for and uh start uh going through the the SEO. But that was the uh the way that we did it in Claude. definitely ran into some bumps and some lessons learned. I think the the way I'll probably extend it a little bit further is maybe do like a a bottom funnel. So if you notice previously I have like the four and the how to use maybe do like a verse where you have like one verse the other uh as far as um either the tech the the local areas or even go after like surrounding areas of the cities to just see if I can pull in more pages. Real quick everyone, if you haven't already, please remember to like and subscribe. It helps more than you know. If you're looking for on-site training, please click the Vibe Coding Retreat down below. And with that, let's get back to it. So what I'm doing now is I'm actually going to read and then analyze all the pieces and do a bottom up uh keyword and as it's going through we're identifying the extensible architecture and implementing this. So what that means is all this the keyword targeting matrix and things like that. Uh I'm going to go ahead and leverage and try and do um I'm probably going to add a directory after this as well. So I'm going to go ahead and just let this run uh and see how it can implement this. Uh, so we just saw a pop from that and let's edit. So, it's adding to the Okay, cool. So, yeah, we do want to make that we'll say yes, don't ask again. So, it's going ahead and adding to the uh programmatic SEO. Uh we're making sure that it's extensible. And then after that we'll um take a look at uh making a directory. So now we're going to say make uh directory. He did. All right. So, let's see what this SEO directory is. Uh, and if it's uh anything cool. Hopefully we don't crash or we can get like a total number or something. Whoa that is crazy. Uh, all right. This is like way interesting. So uh it gave the bottomup funnels created used to create AI chatbot. Um I was not expecting 2,000 uh total pages from the bottle up longtail funnel. That's uh pretty awesome. Um, and we can we can take this and change it, right? Like this is the cool part about it is we all we need to do is from I think like one, two, three, four, five, six. So five or six base pages. We then created 2,000 pages. Uh that's just that's that's wild. So definitely going to expand Claude uh expand this with more keyword research, fine-tune this better. Uh but I I'm happy with this as a first pass. All right, that's it for us today, everyone. So what we went through was the ability to leverage cloud code and actually take the programmatic SEO system and actually expand on it. So using different types of MCPS, we could actually do a keyword analysis and actually expand the data to a much higher range. With that, happy nerding.

---

*Generated for LLM consumption from nerding.io video library*