The pipeline

How it works

One pipeline runs every article from idea to indexed — with a gate at the point where most automated content goes wrong.

1. 1

   Research

   Every article starts from search data, not a topic list.

   • Pulls keywords and questions your competitors haven’t answered well
   • Checks intent and difficulty before a topic is queued
   • Gathers the facts, figures, and primary sources the draft will cite
2. 2

   Write

   A draft in your site’s voice, structured to rank and to read.

   • Follows per-site content rules: tone, reading level, allowed claims
   • Rotates format (guide, comparison, listicle, explainer) so nothing reads templated
   • Adds internal links and a clear, single search intent per piece
3. 3

   Quality gate

   Nothing queues until it passes automated review.

   • Flags unverifiable claims, wrong prices, and competitor mentions
   • Verifies product facts against your source of truth
   • Blocks near-duplicates by title, target keyword, and structure
4. 4

   Image

   A hero image per post, matched to your brand.

   • Generated to your palette and style, not stock
   • Carries an AI-disclosure caption
   • Sized and compressed for fast loads
5. 5

   Publish

   Released the way a person would, not a script.

   • Drip-published across the day — never a batch dump
   • Randomised times, so there’s no visible cron pattern
   • Daily volume ramps slowly instead of spiking
6. 6

   Monitor

   A living system that watches its own output.

   • Sets a daily traffic baseline from Search Console
   • A significant, sustained drop auto-pauses publishing
   • Pauses are reported with the reason, so you decide when to resume