Full Site Editing Arrives in WordPress 5.9: Adopt Deliberately

WordPress 5.9 shipped this week, and with it the Gutenberg project’s founding promise is structurally complete. Full site editing is in core. Headers, footers, page templates, and the site’s global design surface are now editable in the same block paradigm that reached post content in 2018 and widget areas last summer. It is governed by the global-styles system that grew from theme.json, and delivered through a new generation of block themes built for the architecture.

Four years and change from a controversial editor to a coherent platform. We wrote the beta advice in 2017, reviewed the 5.0 launch honestly, and marked last year’s cadence lessons. So we get to mark the arrival with earned perspective. The destination looks like the component-based, token-governed architecture we have argued for across every layer it covers, now standing complete at the web’s largest install base.

The arrival being structural does not make it universal, though. The useful question is who should adopt the new capability now, who should wait, and what control it shifts to content teams. Who should do what, when, and what the new capability redistributes. Start with adoption by situation, then look at the more lasting question: which controls should move from developers to content teams, and which should stay governed.

Adoption advice, by situation

Existing sites: no action, and mean it. Classic themes remain fully supported with no deadline announced or implied. Classic themes keep getting improvements. Nothing about 5.9 converts a working site into a migration candidate. The upgrade itself is routine, the standard staging pass applies, and it changes nothing about how a classic-theme site operates. Treat FSE’s arrival as a milestone to note, not a project to fund.

New builds: a real look, for some. The global-styles system as a governance mechanism has real appeal, and we mean that structurally. Design constraints enforced through configuration. Brand tokens governing every template. The drift-prevention that design-system thinking has wanted since the customizer era.

Against that: first-generation block themes remain pioneer territory. The ecosystem is young and the patterns are still settling. The accessibility and translation edges are getting exactly the community scrutiny the 2018 experience taught everyone to apply early. That argues for selective adoption. The right new build for FSE this year has a team comfortable on the frontier, requirements the block-theme model serves cleanly, and no dependence on the mature-theme ecosystem’s long tail. Our own new-build defaults move to hybrid setups this year: classic architecture with theme.json governance, block themes where the pilot criteria hold. We expect the default to flip as the ecosystem matures, on the platform’s now-demonstrated cadence.

The redistribution, which is the durable subject

• What FSE actually moves: page and template control shifts from the developer’s deploy pipeline toward the site owner’s editing screen. The header adjusted without a ticket. The template composed without a theme release. This is the ticket-queue pathology, the oldest complaint in the CMS files, being answered at the architecture layer, and its benefits are real and immediate for the people who have waited for them.

• What it therefore reopens: the freedom-versus-guardrails bargain every CMS deployment runs. Every control moved to self-service moves a category of risk with it, and the brand consistency, accessibility, and structural health of the site now depend on which controls are exposed to whom.

• The setup duty: 5.9’s defaults open things up widely, and the careful deployment locks things on purpose. Global styles governed by the design system’s tokens. Template editing scoped by role. The block palette curated per the expose-and-lock rule. And the decisions written down as policy rather than inherited as defaults.

What this changes for agencies

Full site editing repeats a familiar CMS tradeoff. More control moves toward content teams, which can reduce ticket queues and speed up publishing. Without guardrails, the same flexibility creates design drift and accessibility problems. The deciding factor is how deliberately the controls are configured.

Our role through these changes has remained consistent: help clients decide which controls belong in the editor and which should stay governed by the design system. The launch conversation that decides, explicitly and in writing, which surfaces are open: the campaign pages the marketing team should own outright. Which are governed: the templates that carry brand and accessibility duties, editable within token constraints. And which are locked: the structural tier where changes are deployments, whatever the interface offers. Then the role-mapping that puts each surface in the right hands, and the review cadence that adjusts the perimeter as teams demonstrate fluency.

Brand safety and editorial speed, both by design, neither by default. The sentence has survived every revolution of this wheel because both failure modes remain fully available at every capability level. The locked-down site that recreates the ticket queue. The open site that dissolves its own system. The platforms, having now shipped the capability, have done their half. The other half was always governance, and it still ships only by meeting.

Five years, start to finish

A closing note. From the 2017 beta advice (‘the paradigm is right, the friction is real’) through the 5.0 review and the cadence lessons of 5.8, and now structural completion, the Gutenberg story has been the longest-running live demonstration of how platforms turn. Noisily, in public, with real costs honestly logged. And, when the stewardship holds, toward architectures worth arriving at.

The component model won, as it won in the design systems and every adjacent layer. The transition process got measurably better under community pressure. The sites that followed the advice through the turn crossed without drama, which was always the product being sold here. Evaluate early. Deploy deliberately. Bridge with horizons, and train like it matters.

What remains is maturation coverage. The block-theme ecosystem’s growth, the governance patterns settling into practice, the eventual default-flip for new builds. The advice travels intact. Platforms this large turn slowly and in public. The turn is now complete at the structural layer. The capability is ready to evaluate, but it still needs configuration and governance. Adopt deliberately, train the people who will use it, and expect the block-theme ecosystem to keep maturing.