What changes in a marketing organisation when AI becomes central?
Tasks and responsibilities can change as AI takes on more analysis, production and coordination. The organisational effect extends beyond adding a tool to an existing role.
A process may require different handovers, shared information or approval points. Some responsibilities can move closer to the business decision, while specialist support remains necessary elsewhere. These changes do not imply that every team becomes smaller.
BCG’s June 2026 study of 300 global CMOs describes more advanced teams reorganising around business objectives or customer segments. It is a reported pattern in that research, not proof of the right structure for every company. The picture may change as tools develop. BCG’s research
Which marketing roles are affected by AI?
Roles containing research, content production, analysis or coordination can be affected, but task automation does not map neatly onto whole jobs.
A role also includes context, judgement, relationships and accountability. Reducing time on one task may create capacity for other work rather than remove the need for the role.
The effect depends on what the organisation chooses to change. A team using AI within existing processes will face different responsibilities from one redesigning how information, work and approvals move across the function. Skills in directing and evaluating AI become relevant alongside the existing marketing disciplines.
What can sit centrally and what can remain in local teams?
Shared standards and infrastructure can sit centrally while teams retain responsibility for work requiring local knowledge. The balance depends on the organisation and its markets.
Central responsibilities may include approved tools, access controls and common brand information. Local teams may contribute audience evidence, language and knowledge of commercial conditions. Claims or regulated activity can create additional approval responsibilities.
A centre of excellence is one organisational option for providing expertise. An AI council is another mechanism for coordinating decisions. The existence of either does not establish that capability has reached the teams doing the work.
How can the effect of a new team structure be assessed?
Its effect can be assessed through the quality, speed, cost and accountability of the work it supports.
Headcount alone cannot show whether a structure works. Fewer handovers may help one process while reducing specialist input in another. Time released by automation also has different value depending on how it is used.
Evidence from completed work can reveal where decisions remain unclear or effort has shifted rather than disappeared. That gives a more specific account of organisational performance than adoption figures or the number of AI tools available.
Legal note: This answer provides general information, not legal advice. Seek advice from qualified legal counsel for your circumstances.