AI Content Automation

The problem

Most businesses know they should publish content. Almost none of them have time to do it well. The result is either an empty blog, an expensive agency, or AI slop pasted straight from a chatbot — thin articles that read like everyone else’s and rank for nothing.

The interesting problem isn’t “generate text”. It’s building a repeatable editorial system: research what’s worth writing, write it in the client’s voice, keep facts straight, link it into the rest of the site, and publish on schedule — without a human doing the boring parts by hand.

What I built

A content pipeline on n8n that runs the whole loop:

1. Research — the pipeline gathers what the article needs before a single sentence is written: topic angles, questions people actually ask, what competitors already cover.
2. Writing — drafts are generated against a strict brief: structure, tone-of-voice rules, terminology the client uses, and things it must never claim.
3. Editorial checks — internal linking, formatting, and a review gate. A human can approve, edit, or reject before anything ships.
4. Publishing — the approved article lands in the client’s CMS on schedule, formatted and ready.

Every client gets the pipeline wired into their own stack and tuned to their niche — the system is the same, the voice never is.

What a typical engagement delivers

Across current clients, the same pattern keeps repeating:

• +12% more inbound inquiries within the first three months — content that answers real questions brings people who ask for quotes,
• +42% site visibility, and climbing — more indexed pages showing up for more of the queries your customers actually type,
• break-even around month two on monthly retainers — from then on the pipeline costs less than the traffic and leads it brings in.

No guaranteed rankings, no magic. These are typical numbers, not promises — starting points and niches differ, and anyone who tells you otherwise is selling you the hype I filter out of articles.

Why it doesn’t read like AI slop

Here’s the part most automation vendors can’t say: I’m a copywriter first. I can write your blog by hand — I’ve done it for a living. The pipeline exists because I took my own writing process apart and turned it into rules: how an article should be structured, what a claim needs behind it, where the tone slips into hype, what makes a paragraph earn its place.

So the AI isn’t “generating content”. It’s executing my craft at scale — and I stay in the loop as the editor. On managed plans, every article can pass through my hands before it ships: a working copywriter reading it the way a customer will, not a model grading its own homework.

Judgment stays human either way. Positioning, offers, and anything a lawyer would care about is never left to the pipeline — automation removes the repetitive 80% so the human 20% actually happens.

Two ways to work with me

001 — YOUR SYSTEM (handoff). I build the pipeline for you, deploy it on your infrastructure, wire it into your tools, and hand it over — documented, tuned, yours. You run it; I stay available with limited support. 19 of my 26 clients chose this.

002 — MY SYSTEM (managed, monthly retainer). I host everything and operate it for you. Two flavors: full-auto — the system runs and articles get made — or copywriter-checked, where I personally review and polish every article before it goes live. 7 clients run this way today.

What’s included

Both models ship with more than “a workflow”:

• Full integrations — CMS, analytics, and the tools you already use
• WordPress configuration — set up and tuned, not just connected
• Image generation setup — article graphics on autopilot
• Free content plan — the first topics mapped before we start
• Free documentation — the system explained in plain language
• Unlimited reviews — revision rounds aren’t a billable event
• Support — a person answers, and it’s the person who built it
• Monthly performance report — what published, what ranks, what’s next

Why now — the market in 2026

Industry numbers explain why this service exists: 97% of marketers plan to use AI for content in 2026, and 74% of newly published web pages already contain AI-generated content. But only ~2.5% of those pages are pure AI — the ~73% that blend AI with human writing are the ones producing the strongest results.

That blend is exactly what I sell. Everyone has access to the same models; almost nobody pairs them with an actual editorial process and a copywriter who reads the output. The gap between “we use AI” and “our content works” is where this pipeline lives.

Where it runs

Currently in production for businesses in:

• Ecommerce — peptide and sport-supplement brands, where content has to be careful with claims and specific with facts,
• Pest control services — local intent, seasonal topics, service pages that need to answer real questions,
• Travel services — destination and arrangement content with a lot of practical detail.

Boring niches are a feature, not a bug. That’s where consistent, factual content still wins.

The product that fell out of it

The workflow wasn’t born as a product for clients — I built it for my own content first, and it simply worked. Clients came next. At one point I even turned it into a standalone app: Articfly, a self-serve blog article generator. As a SaaS it didn’t find its customers, so I sold it and moved on — the service it came from never stopped running.

That’s the honest shape of this business: the service pays, the products compound, and not every experiment has to work for the system to win.

By the numbers

• 1.5 years in operation
• 26 clients total — 19 pipeline handoffs, 7 monthly retainers
• +12% inquiries in a typical first quarter; +42% site visibility
• Month 2 — the point where a retainer typically pays for itself
• 3 niches in daily production: ecommerce, pest control, travel
• Every retainer article available with a copywriter’s manual pass