Case study · Multi-site content operations

Running a Content Programme Across Six Estates

A content operation covering six separate estates, where the hard problem is not producing articles but stopping defective ones from reaching a live client site.

The situation

Running content across six separate estates, each with its own platform, its own brand voice and its own commercial rules, creates a problem that does not exist on one site: a defect introduced once appears everywhere before anyone notices.

Volume makes manual review impractical, and manual review is unreliable in exactly the way that matters, because the defects that survive are the ones that look correct.

What was done

  • Built a pipeline with a quality gate. Research, drafting, fact checking and review as distinct stages, with an automated compliance gate that blocks publication on defined hard failures rather than warning about them.
  • Encoded the rules that had been informal. Banned vocabulary, opening patterns, heading requirements, source verification and structured-data requirements written as checks rather than as guidance nobody re-reads.
  • Added anti-template rotation. An audit of the published estate showed structural repetition across sites: near-identical FAQ counts, closing headings and section shapes. The pipeline now varies these deliberately per post.
  • Ran a full estate audit. Every published post re-checked against the current standard rather than assumed correct because it passed at the time.

Results

  • 1,315 posts audited across six estates
  • 6 separate sites under one standard
  • 57 posts corrected in the most recent sweep
  • 0 hard failures permitted to publish

Counts are from the most recent estate-wide audit, 27 August 2026. The correction figure is the useful one: 57 posts out of 1,315 needed fixing, which is the rate a gate has to be designed around. A process that assumes published work stays correct will not find them.

What transfers to other businesses

At any volume, the constraint is not production, it is the defect rate. The check that matters is the one run against work already published, because those are the errors that have been live longest and that everyone has stopped looking at.

Related services

Talk about your situation

The first conversation is short and mostly questions. Get in touch and tell me what you are trying to fix.