Services About Results Blog Book a call

How We Built an AI Content Factory That Publishes 30 Posts Per Month

Laptop and notebook representing content creation

Content marketing works — but it's brutally labour-intensive at scale. A single high-quality blog post traditionally takes a skilled writer 4–8 hours to research, write, edit, and publish. At 30 posts per month, that's 120–240 hours of writing time, or roughly 1.5–3 full-time employees just for content.

We built a content factory system for a client that produces 30 polished, SEO-optimised blog posts per month for under £800/month in total costs, including tooling. Here's exactly how it works.

The System Architecture

The content factory has five stages, all automated and orchestrated in n8n:

Stage 1: Keyword & Topic Harvesting

A scheduled workflow runs weekly, pulling keyword data from Ahrefs (via API) for the client's target topics. GPT-4o processes the keyword clusters and generates 40 potential post titles ranked by search volume, difficulty, and business relevance. A human editor reviews this list monthly and approves 30–35 titles, which populate a content calendar Airtable base.

Stage 2: Research Compilation

For each approved title, an n8n workflow automatically performs a Bing Search to gather the top 5 ranking articles on that topic, extracts their key points using a web scraper + GPT summary, and stores structured research notes in Airtable alongside the title.

Stage 3: AI Drafting

When a post's scheduled publish date is 7 days out, the workflow triggers GPT-4o with a carefully engineered prompt that includes: the target keyword, the approved title, the research notes, the brand voice guide, the desired word count (1,200–1,800 words), and instructions to include headers, a strong opening, and a clear CTA section. The draft is written into a Google Doc automatically.

Stage 4: Human Edit Pass

A notification is sent to the editor's Slack with a link to the Google Doc draft. The editor spends 20–30 minutes reviewing, adjusting tone, adding any proprietary insights, and approving. This is the only manual step in the whole pipeline — and it's intentional. Human editorial oversight is what keeps the content from being generic.

Stage 5: Publishing

Once the Google Doc is marked as "Approved" in Airtable, the workflow automatically formats the post for HTML, generates a meta description and Open Graph tags using GPT, uploads a featured image sourced from Unsplash (with the appropriate license), and publishes to WordPress via its REST API — with the correct category, tags, and scheduled publish time.

The Results

Within 6 months of running this system, the client's organic traffic grew from 1,200 to 9,400 monthly visitors. Their domain authority increased from 18 to 31. They ranked on page 1 for 47 new keywords.

Total monthly cost of the system: £780 (tooling + editor time). The equivalent output from a traditional content agency would cost £4,500–£6,000/month.

"We went from publishing 2 posts a month to 30, without adding headcount. The quality is actually higher because we have more time to focus on the editorial review." — Head of Marketing, B2B SaaS

Can This Work for Your Business?

This architecture works best for companies with a clear ICP and topic space, a willingness to invest in an initial 4–6 week setup, and at least one person who can do a 20–30 minute editorial review. Businesses in professional services, SaaS, and B2B technology are ideal candidates.

Ready to publish more content without more headcount?

Systro builds custom AI content factories for B2B companies. We handle the full setup — keyword strategy, AI drafting, CMS integration — so you can focus on approvals and growth.

Book a free call Email us directly