My SEO Approach Using Typemill and AI

While I’m not an SEO expert, I’ve spent over two decades working with websites and content, including my ongoing work on the Typemill project. Last year, while exploring different marketing strategies for Typemill, I decided to focus on SEO for a while. In this article, I’ll share what I learned, and how I replaced some external tools with native Typemill features. By the end, I’ll also share a few personal conclusions that might surprise you. Let’s dive in.

How I Tried SEO with External Tools

By late 2024, I was experimenting with different marketing strategies for Typemill. Big media outlets mostly ignored it, the newsletter fell flat, and social media brought little traction. So I turned to SEO and hired a professional to improve existing landing pages. We quickly realized these pages had no realistic chance of ranking well, so we pivoted to blog posts focusing on broader, less branded topics.

I launched a new section on the Typemill website and started using an expensive, AI-powered SEO service that promised fully optimized articles. The platform handled everything: topic clusters, keyword research, and article suggestions. It even auto-generated well-structured, keyword-rich content that matched the quality of top-ranking posts.

Here's my honest assessment: I put in minimal effort, simply generating articles on interesting Typemill use cases and publishing them as-is. The results were mixed:

• Articles had solid structure and looked high-quality
• But the content felt generic
• Rankings were mediocre with zero conversions
• However, the ranking data provided valuable insights

After a few weeks, I pulled the plug. Generic AI articles just weren’t worth the investment.

My Switch to Typemill's Built-in Tools

After launching AI integration for Typemill in March 2025 and enhancing the SEO plugin, I was finally able to build a full workflow for writing and optimizing articles without relying on third-party SEO tools. While it doesn’t offer the feature depth of premium SEO platforms, it covers everything I actually need.

Let’s start with the SEO plugin. The setup can be a bit tricky for newcomers, since it involves Google’s complex authentication process. But once it’s configured, you get direct access to Google Search Console data for each page right inside Typemill:

Sure, you can access all this data in Google Search Console. But having it right next to your content, already filtered and focused, makes things much easier. It also opens the door to workflows that directly integrate this data into your content optimization process.

Here’s my personal workflow: After setting filters for country, start date, and end date, I scroll through the list of search queries to find those relevant to my page or product. Sometimes I’ll check how a query has evolved over time, look at which countries it ranks well in, or open it in Google to review the actual search results. If a query looks promising, I mark it using the checkboxes on the left.

You can also use filters to surface relevant queries more efficiently:

• Top positions: Queries ranking 1-10 (first page results)
• Hot chances: Queries ranking 11-19 that could reach page one with optimization
• High traffic: Queries with lots of impressions, even if rankings are low

Once I’ve marked promising queries, I filter them for my final list.

Leveraging AI for Deep Analysis

Below the page filters, there’s a button to copy query data from the current filter as a Markdown table. I paste this table directly below my article for easy reference. I also bring in data from my external analytics tool, Matomo. I copy Matomo data into ChatGPT or Claude and ask for a well-formatted Markdown table, which I then paste back into the article. This way, the bottom of my articles now includes detailed insights:

Next, I open the article in Typemill's AI interface and run analysis prompts.

My current prompt is fairly generic and definitely needs improvement. Still, the results from Claude have been quite helpful for me:

You can try it yourself and store different prompts for SEO analysis in the prompt library. Typemill creates a new article version for each AI response, so you can easily compare results using pagination.

You can also create flexible workflows by crafting several follow-up prompts and applying them step by step. For example, you might run separate analyses for E-E-A-T or keyword targeting, then, once you identify valuable insights, ask the AI to apply those suggestions directly to the article. With some experimentation, you’ll discover prompts and workflows that deliver the best results.

When you’re satisfied with your article, head back to the SEO plugin and open the Links tab. Here, you can view internal backlinks pointing to your article or search for relevant articles on your website to create new backlinks. This feature requires the BetterSearch plugin, so be sure to install it.

Again, these workflows don’t offer the power of external SEO tools. But for me, they provide plenty of flexibility without breaking the bank.

Building a Comprehensive Website Strategy

Analyzing and improving individual pages is important, but it’s only part of the picture. By providing AI with relevant data, you can review your overall website strategy and get assistance with strategic decision-making. For this kind of analysis, I don’t use Typemill directly. Instead, I switch to services like ChatGPT or Claude. Here’s an example prompt I started with:

Can you check the following traffic analysis for Typemill. I have several goals but the most important are: 

* Download the Typemill zip (indicates that users install it)
* Visit the license page
* Click on a buy button

I evaluated traffic with goal Download and click on buy button. 

Interestingly all my SEO content from knowledge-hub does not seem to work well. For me it looks that traffic from other related pages and directories, and traffic from reddit and github work well. Traffic from search engines is unclear, but since SEO content does not perform, I guess these are mainly product searches for Typemill that convert. 

What do you think and do you need more data for analysis?

## Evaluation 01.01.2025 - 31.06.2025

* Total Visits: 13.500
* Total Downloads: 850
* Visits on License page: 1.500
* Click on Buy: 22

### Goal Download

| Source            | Visits | Page Views | Pages/Visit | Avg. Duration | Bounce Rate |
|-------------------|--------|------------|-------------|----------------|--------------|
| Search Engines    | 332    | 3,895      | 11.7        | 14 min 14s     | 0%           |
| Websites          | 262    | 3,428      | 13.1        | 14 min 54s     | 0%           |
| Direct Entry      | 171    | 1,337      | 7.8         | 5 min 58s      | 2%           |
| Social Networks   | 81     | 727        | 9           | 7 min 48s      | 1%           |
| Campaigns         | 4      | 25         | 6.3         | 3 min 0s       |              |

| Referrer                        | Visits |
|--------------------------------|--------|
| cmsstash.de                    | 41     |
| reddit                    | 37     |
| github                    | 36     |
| try.typemill.net               | 30     |
| www.XXXXXXXXXXXX.io            | 24     |
| trendschau.net                 | 18     |
| localhost                      | 15     |
| XXXXXXXXXXXX.org               | 8      |
| XXXXXXXXXXXXXXX.info           | 6      |
| XXXXXXXXXXXXXXXXXXXXX.de       | 5      |
| chatgpt.com                    | 4      |

### Entry Point Knowledge Hub

* Visits: 1800
* Downloads: 33
* Buy: 1

During the discussion, I gradually added more data for deeper analysis and finally asked ChatGPT to summarize all findings in Markdown. Since I also use Typemill to manage larger notes, editorial plans, and similar content, I created a new (unpublished) folder called “Analysis” and stored the summary there. Next, I requested recommendations based on the very limited resources I have. Now I have a strategy for the next half year.

My Final Take on SEO

The insights from strategic analysis are often quite straightforward. Sometimes you don’t even need an AI chatbot to see them. That said, I find these chatbots incredibly helpful for organizing my thoughts and exploring ideas.

Listicles, how-to guides, and best-of articles might still work well for some projects, but for Typemill — especially in this AI-driven age — they don’t feel like the right fit. That kind of content is everywhere now, often heavily keyword-optimized and quickly generated by AI. While it can sometimes rank decently, it rarely drives meaningful conversions or sparks genuine interest. Plus, it’s just not very enjoyable to create.

That doesn’t mean SEO isn’t valuable. I absolutely use SEO tools and workflows to support my content. But after all this experimentation, I’ve decided to focus less on SEO and more on writing personal stories and sharing my own experiences with Typemill. I don’t start with SEO analysis or keyword research. Instead, I write about what genuinely interests me and what I think readers might find relevant.

There’s a name for this approach in SEO circles: E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness). For me, though, it’s simply blogging the way it used to be: authentic, straightforward, and personal. That’s how it all started, and it’s what I still enjoy the most.