Architecting Contentstack's content model

Problem

AEO's legacy CMS, Oracle WCS, had two structural issues that compounded each other: content was tightly coupled to its visual presentation, and reusable elements were scattered (and often duplicated) across the system. Moving to a headless CMS — Contentstack — was the opportunity to fix both, but only if the new architecture was designed deliberately.

Approach

After extensive discovery with the design and content authoring teams, I anchored the architecture on three principles:

1. The "lockup" as the atomic unit. A single marketing message lives in one place — the lockup — and is referenced wherever it needs to appear. The message is authored once and consumed everywhere, ensuring consistency across surfaces.
2. Visual control abstracted from the message. Each surface (web, mobile, app) can style the lockup differently without changing the underlying content. Brand voice stays consistent; visual treatment adapts to context.
3. A shared content library. Reusable entries live in a discoverable library, eliminating the redundancy and bloat that plagued the WCS era.

Outcome

• Smooth transition from Oracle WCS to Contentstack
• Significantly less content duplication across channels
• Cohesive brand message across web, mobile, and app surfaces
• Faster authoring — and easier audits — because every message has one canonical home

What this says about how I architect

The right abstraction is the one that lets non-engineers move fast without breaking things. The lockup model wasn't novel — but pairing it with abstracted visual control and a real content library is what made the system actually maintainable as content volume grew. Architecture decisions live or die on the day-to-day experience of the people authoring against them.