There is no model that is absolutely correct, but there are models that are frequently wrong.
Most healthy organizations ultimately operate in a hybrid model: a core team dedicated to maintaining stability, alongside contributors embedded within product teams. If you are in the early stages, start with a focused approach; definitely do not begin with a federated model, as lacking a central core will dilute any model. Choose a model based on actual team size and budget, not industry trends.
Three management models for the design system are often mentioned: centralized (a core team dedicated to creation and delivery), federated (representatives from multiple product teams collaborating), and hybrid (a combination of both). The practical question is not "which model is best" but "which model fits your current scale and budget."
The centralized model places all responsibility on a dedicated core team. This team builds components (interface elements), maintains standards, and distributes them for product teams to use. The advantage is high consistency and quick decision-making. The limitation is that when the organization grows rapidly, the core team becomes a bottleneck, unable to meet all demands.
The federated model decentralizes contributions from representatives of product teams. It sounds democratic and cost-effective, but according to Nathan Curtis in The Fallacy of Federated Design Systems, federated is not the first choice and will never run effectively without central investment. This is not an independent model but an extension layer placed on top of a solid centralized foundation.
The hybrid model maintains a fixed core team to manage the foundation and common standards, while having contributors embedded in product teams to bring real needs into the system. This is the goal of most healthy systems in the maturity stage.
Choosing to focus when: the organization is just starting to build a system, needs a quick and highly consistent foundation, and has few shared teams. This is not a poor model; it is the right starting point.
Transition to hybrid when: there is a solid enough foundation, team size and product count increase, and the core team starts to struggle with the pace of demands. At this point, integrating contributors into each product team is a natural way to expand, not to discard the core.
Purely federated is not a reasonable choice at any stage if the definition is understood correctly: without a central core, no one owns the standards, and the system will disintegrate over time.
Federated is not an option but an aspect. It should never be pursued first and cannot function without central investment.
Nathan Curtis, The Fallacy of Federated Design Systems (EightShapes)
In projects, Sinh Vũ often establishes a focused foundation during the implementation phase, with the Sinh Vũ team playing a temporary core role. Upon handover, the first question Sinh Vũ asks the client team is: internally, who will own this system? If there is no clear answer to that question, then discussions about hybrid or federal models should not take place, as a lack of a central core makes any structure diluted.
The common roadmap is: starting with focus, stabilizing the foundation, identifying who is core within the client's organization, and then gradually opening up contribution mechanisms from product teams, which means moving towards hybrid. The speed of this transition depends on the number of teams, operational discipline, and the budget to maintain the core; there is no one-size-fits-all formula for every organization. This is something that should be discussed directly based on your specific situation.
Topic: Choosing a design system governance model: centralized, federal, or hybrid. Sinh Vũ guide, sinhvu.com
Select each item you find appropriate, then print or save as PDF to take with you.
If you have marked most of the signs above, this is the time to discuss in more detail. Sinh Vũ can help you review and propose a direction.
Nathan Curtis, Team Models for Scaling a Design System (EightShapes). Nathan Curtis, The Fallacy of Federated Design Systems (EightShapes).
No. This is the most common misconception. Federation is a way to expand contributions from product teams, not a replacement for the core. If the core is removed, the system will have no ownership and will gradually die. Federation only works when there is a strong enough centralized foundation.
At a small scale, the more important question is who owns and who makes the final decision; it is not necessary to name the model. In reality, most small companies operate in a naturally centralized manner because only a few people are involved. When expanding the team, clearer thinking about contribution mechanisms and governance is needed.