One Source of Truth, Long Before AI Made It Easy

On paper it was one section. In practice it was six or seven operations pretending to be one: images in one system, news copy in another, sponsor materials over email, advertising with a different team on a different schedule, production planning on its own spreadsheets, financial reporting reconciled separately after the fact. Nobody designed it this way; it accumulated dozens of reasonable workarounds that never talked to each other.

One central, publisher-facing system that every function fed into and drew from: images, editorial, sponsors, advertising, production schedules, financial reporting, all visible from one place instead of scattered across six. Editorial could see what advertising had committed to; production could see what finance was reconciling against. For the first time, the section had a single version of the truth instead of six competing ones.

There's no number to put on what that was worth; the figures don't survive twenty years later, and this isn't the place to invent them. What can be said plainly: the operation ran more smoothly, with far less time lost reconciling versions of the truth that should never have diverged, and it became the standard approach used across the rest of the group.

The only way through was mapping where information lived and agreeing which version was authoritative. AI hasn't removed that groundwork, it's raised the cost of skipping it. Zalando's engineering team has since described hitting the same wall at far larger scale, fixing it the same way: govern the source of truth first, then layer AI on top.

That's the discipline behind Solus Command's Governance Register: dedicated function workspaces reporting up to the one picture this newsroom only got by hand.