Mave: enterprise content without the enterprise drag

I led Mave from positioning through editor adoption — a MACH-based headless CMS for enterprise marketing teams drowning in monolithic platforms. My job was to make content delivery composable without making editors feel like they were using developer tooling. Enterprise CMS projects often fail quietly: IT gets composability, marketing gets frustration. I named editor satisfaction as a first-class metric alongside deployment speed and ops cost — if editors hate it, the migration was wasted.

Marketing wanted to ship pages and campaigns faster; IT wanted control, auditability, and lower infrastructure cost. Editors were stuck in workflows that required tickets for changes that should take minutes. Every release cycle burned goodwill. Teams quoted "headless" as the future but still operated like the old CMS was haunting the building. The trap was building for architecture slides instead of the person publishing at 4:55 PM before a launch.

I bet on MACH architecture — microservices, API-first, cloud-native, headless — but only where it moved measurable outcomes: deployment speed, ops cost, and editor satisfaction. The product wedge was "composable enterprise content" with an editor experience that didn't require a ticket to IT. Composable where it reduced drag; simple where editors lived daily.

The fight was scope. Stakeholders wanted every enterprise feature on day one — workflows, permissions, localization, DAM hooks, personalization. I sequenced: core content modeling and delivery first, then composable integrations on a stable base. Bi-weekly releases with editor feedback loops beat a big-bang launch nobody would adopt. I killed features that demoed well but didn't move deployment time or editor NPS.

Deployment cycles dropped 30%. Operational costs fell 20%. Content delivery speed improved 40%. Editor satisfaction hit 95% — the metric I cared about most. A technically correct headless CMS that editors hate doesn't ship twice. Marketing teams started owning updates again instead of queueing them.

I'd bring editors into roadmap reviews earlier — not as "users" but as veto power on anything that added clicks. Their patience is the real SLA. I'd also publish internal scorecards tying every epic to deployment time, ops cost, or editor satisfaction. Without that, headless becomes an architecture science project.