The Pugmill Trifecta

In a ceramics studio, a pugmill transforms trimming and slop into perfectly usable clay. The machine forces the reclaimable material through a tapered barrel using an auger screw, compressing, de-airing, and wedging the clay, making it ready to throw pots on the wheel.

SEO-optimized content is like old clay and needs to be transformed into Answer Engine-ready content. A content management system processes these inputs into publishable outputs. The Pugmill set of projects applies the mechanical processing model to digital content across three distinct software components: a standalone rebuildable CMS, a WordPress plugin, and an analytics network.

1. Pugmill CMS - Open Source Thought Leadership

Pugmill CMS is an open-source platform built on Next.js, PostgreSQL, and Drizzle ORM. It provides a headless-ready architecture with an integrated front-end theme system. The platform includes a visual editor, pluggable storage options for local or S3-compatible environments, and Answer Engine Optimization metadata fields. The system generates RSS feeds, sitemaps, and llms.txt files to supply structured data to crawlers.

2. AEO Pugmill Plugin for WordPress - Network Data Collection

Answer Engine Optimization (AEO) shapes how AI-powered tools like ChatGPT, Perplexity, and Google's AI Overviews pull information from your site and surface it in direct responses. Where traditional search engine optimization targets ranked links, AEO targets the cited answer itself. The signals that matter shift accordingly. Instead of keyword density and backlink volume, the signals that matter are clear entity definitions, structured data markup, authoritative sourcing, and prose that directly answers a specific question. The components described in the following sections each address one or more of those signals. The AEO Pugmill plugin structures data for existing WordPress installations to accommodate AI crawlers. It generates machine-readable endpoints, including llms.txt files and structured Markdown renderings of individual posts. The plugin injects FAQPage schema, entity mentions with reference URLs, and citation JSON-LD into the page header. Supplying structured data gives crawlers like GPTBot, ClaudeBot, and PerplexityBot parse-ready facts and Q&A pairs to extract.

3. The Pugmill Analytics Network - Data Aggregator

AI crawlers and search bots index web content. The AEO Pugmill plugin logs incoming requests from these bots. It records the canonical bot name, the specific resource type requested, and the date. The system distinguishes between standard HTML page visits and requests for structured formats like Markdown or JSON-LD.

Sites choosing to participate in the Pugmill Intelligence Network send daily aggregated traffic summaries into a shared pool, where bot-signature patterns identified across thousands of installations improve the filters applied to each individual site's data. The result feeds directly into the content metrics that editors and strategists rely on: session counts, scroll depth, and conversion attribution tend to be more accurate because non-human requests are removed before those numbers reach the editorial dashboard. Content decisions made from that cleaned data, whether about topic prioritization, publishing cadence, or audience segmentation, reflect reader behavior with crawler and scraper traffic The pipeline excludes bots and scrapers before passing counts to the Pugmill dashboard, keeping the engagement figures tied to real reader behavior. intelligence network. The central dashboard compiles this data to show which bots request specific content formats and how crawling behavior shifts over time. Tracking these requests creates a public record of how AI systems interact with structured web content.

Business Strategy

As this three-part product suite gains traction, its value will grow in the data it collects. Pugmill CMS and the AEO Pugmill plugin will send data to the AEO Pugmill Intelligence Network. The data at the end of this process is the true value.

I will transition the Pugmill Trifecta from the Janzen Works venture studio umbrella into its own Delaware C Corporation, along with all intellectual property, assets, and data. At this point, it graduates from a fun experiment to a valuable business asset that some key players may be interested in acquiring.

So yes, before I wrote a single line of code, I had a deep understanding of the problem, a carefully scoped solution, a go-to-market strategy, and an existing strategy.

I can do this because I can build software myself for virtually free, not counting my time. The full time required to build this three-part software set has been about six weeks. It will take much more time to gain traction and prove whether there is an exit. If no exit, I have no worries. It cost me little, and I have lots of other products planned. Between that work and other startup founders who want to hire a fractional CPO/CTO like me to build their vision.