2025 reflections on modern Salesforce architecture, exploring how legacy structure, policy-driven design, and execution shape AI readiness.

Most Salesforce estates are not broken because they customised too much. They are broken because the architecture is still org-centric. Data 360 changes the centre of gravity. The org becomes a container, not the system.

Salesforce architects increasingly face the challenge of deciding what belongs inside an application org and what should be modelled in Data 360. This post explains how the two layers work together, where enterprise meaning should live, and how clear responsibility boundaries create cleaner, more scalable architectures.

Enterprises are beginning to explore new patterns for structuring their Salesforce estates. Data Spaces offer a way to govern and segment data without creating new application orgs, while Data Model Objects provide a shared definition of meaning across systems. Orgs remain essential for application logic, but the centre of gravity is starting to shift toward the data layer. This article shares early insights from conversations with leaders, architects, and engineers who are experimenting with these ideas and shaping the next generation of Salesforce architecture.

Interoperability is now the centre of gravity in Salesforce architecture. Modern estates span multiple clouds, data platforms, and external systems. The question is no longer whether to integrate. It is how to design the fabric that keeps data consistent, useful, and trustworthy across the enterprise.

The sObject still matters, but the architectural truth now lives in Data 360. DMOs unify data across orgs, redefining Salesforce as a networked, AI-ready enterprise platform.

Dreamforce 2025 signalled a new phase for Salesforce. This article explores why the deeply integrated platform positions Salesforce to lead in enterprise AI.