Crucially, those shared standards do not dictate how a desk runs its business. The goal is to give each business unit room to determine where true nuance and intricacy live. Doing so shifts the conversation to tangible questions around which definitions must be shared so data can move cleanly across regions, asset classes, and control functions without re-implementing the same basics repeatedly.

Governance: solving the “mayor” problem

Practically speaking, total convergence and lockstep alignment are a nirvana that most firms will never reach, especially when it means people must invest in change that does not seem to benefit them directly. When people believe what they already have works, when they look through the lens of “we have always done it this way,” or when they feel a deep attachment to technology they built themselves, change is that much harder.

Chief transformation officers need three things to build credibility and soften these forms of resistance. First, they need a clear plan and business case. Second, they need a top-down mandate. And third, they need to recognize that mandates from the top still need local buy-in to succeed, across senior management and business units.

In addition, transformation does not end at a particular milestone. Markets evolve, products change, and whatever looks complete in a three-year plan will be overtaken by new requirements, new asset classes, and new operating demands. To succeed, adaptability matters as much as architecture to stay current, absorb change, and keep the firm coherent as strategies shift.

With a layer that creates standards on top of existing systems as they age and are replaced naturally, core activities like regulatory reporting, funding decisions, compliance oversight, management reporting, and client reporting can draw from a consistent set of normalized data. They do not need to be rebuilt around system-by-system definitions or follow every upgrade. For example, if the firm replaces a trading system, the work is to map the new system’s outputs into the ecosystem once, rather than remapping those outputs into every downstream report, control, and workflow.

Once definitions are shared and data is normalized, the focus of transformation can come back around to the machinery that moves data across regions and platforms, enforces runtime controls, and delivers it fast enough to be operationally meaningful. That’s where architecture determines outcomes rather than debate about location or build versus buy.