AI Blog Post Generator for SEO & AI Search

Daily blog posts in your brand voice, optimized to rank in Google AND get cited by AI search (ChatGPT, Google AI Overviews, Perplexity). Researched against authority sources, sitemap-aware to avoid duplicates, ready to paste into WordPress, Webflow, Ghost, or Notion.

Created by Chris Moen • Version 40 • 38 steps

Use this app

AI Blog Post Generator

Get a publishable blog post in your brand voice every weekday morning. Researched against real authority sources. Ready to paste into WordPress, Webflow, Ghost, Notion — anywhere.

What you tell us

Just two required fields at install:

• Your website URL — we crawl your homepage and about page once to learn your brand name, niche, audience, and tone automatically.
• Where to email drafts — any inbox.

Optional: schedule cadence, language, market region, CTA copy, manual brand-name / niche / audience / tone overrides (only needed if your site blocks scraping).

What you get

Every weekday at 8am UTC, a new draft lands in your inbox:

• 1,500–2,000 words in your brand voice
• Question-shaped H2s matching real user search queries
• AEO-extractable answer block in the opening 100 words
• One comparison HTML table — extracts cleanly into AI Overviews
• 3–6 outbound @@INLINE_0@@ links to authoritative sources (.gov / .edu / industry leaders), all from real scraped corpus — never fabricated
• 6–8 numeric / citable claims with named sources
• FAQ section with 4–6 self-contained 40–60 word answers (FAQPage schema-ready)
• Sitemap-aware deduplication — we pull your existing post URLs from your sitemap to avoid generating duplicates
• Auto-derived CTA in your voice (or your override)

What you bring

Nothing. The flow author covers OpenAI, SerpAPI, Firecrawl, and Resend email costs.

Manually trigger or let the schedule run.

Auto-publish to your CMS (optional)

If you have a CMS with a REST API that accepts Bearer-token auth, drafts can be auto-published instead of just emailed. Three setup fields:

• Publish connection — bind any HTTP connection (Bearer auth recommended) whose base-url points to your CMS host.
• Publish path — relative path appended to your connection's base-url, e.g. /wp-json/wp/v2/posts.
• JSON payload template — what your CMS expects. Use {{title}} {{slug}} {{html}} {{excerpt}} placeholders. Defaults to {title, slug, content, excerpt} if blank.
• Extra headers (optional) — non-auth headers your CMS requires.

Template examples

WordPress (with Application Password):

{"title": "{{title}}", "content": "{{html}}", "excerpt": "{{excerpt}}", "status": "draft"}

Webflow CMS:

{"fields": {"name": "{{title}}", "slug": "{{slug}}", "post-body": "{{html}}", "_archived": false, "_draft": true}}

Ghost (custom integration with JWT in connection):

{"posts": [{"title": "{{title}}", "html": "{{html}}", "status": "draft"}]}

Generic webhook (Zapier, n8n, Make.com, custom backend) — defaults work, no template needed:

{"title": "{{title}}", "slug": "{{slug}}", "content": "{{html}}", "excerpt": "{{excerpt}}"}

Leave the publish fields blank to receive markdown by email only. The email is sent regardless of whether auto-publish succeeds — your draft is never lost.

How we choose what to write about

You don't have to pick topics. Each weekday morning we run a 3-stage research pipeline before writing:

• Weekly strategy — Once a week, we generate a 5-slot editorial calendar for your niche. Each slot is framed around a real audience question your readers Google late at night, not around product variants. The strategist agent mines Reddit, Quora, and SERP related-searches for actual audience language.
• Topic for today — From the weekly calendar we pick today's weekday slot (Monday's "panic-button" topic differs from Friday's "weekend project" angle). Calendar regenerates as it ages, so topics stay fresh.
• Per-article research — Once the day's topic is locked, two parallel agents run:
• Organic research: SerpAPI on the topic, then Firecrawl-scrape the top non-social results for a corpus of ~4-6 high-quality sources
• Authority research: SerpAPI with site: filters against domains relevant to your niche (.gov, .edu, industry research bodies, recognized trade publications), then Firecrawl-scrape the top 3-5 for citable stats

Every numeric claim, percentage, and named authority in your article comes from this scraped corpus — never fabricated. The writer cites them by name with outbound <a href> links to the original source.

Sitemap-aware deduplication

Before generating a title, we fetch your sitemap (/sitemap.xml auto-detected, or paste a custom URL at setup) and pull all your existing post URLs. The title generator checks each candidate slug against this list and rejects collisions before writing. New articles never duplicate old ones — even after 100+ runs.