Marketing has outgrown the organization built to run it | MarTech

Marketing work spans interconnected systems and teams, yet authority remains divided among specialized functions. That disconnect makes it harder to coordinate decisions, manage accountability, and deliver results. AI is exposing the gaps as more marketing work moves into automated workflows.

In 2011, the first Marketing Technology Landscape contained around 150 products. The 2026 edition contains 15,505. The story isn’t the frankly ridiculous number of products marketers can buy. It’s how that technology has changed the way marketing gets made.

CRM, automation, DAM, personalization, workflow, analytics, commerce platforms, and now AI have connected activities that once operated far more independently. Data moves between systems, content moves from brief to production to activation, and customer behavior can change what happens next. Technology is now part of the infrastructure through which marketing is produced, rather than just supporting marketing activity.

The organizational response is entirely logical. As technology grew more sophisticated, businesses added specialist capabilities around it: marketing operations (MOps), martech, creative operations (CreativeOps), CRM, analytics, ecommerce, and deep expertise in data and technology. But this specialization developed alongside growing interdependence. Expertise sits within defined functions, while the production system spans them.

Specialization met complexity, but interdependence changed the problem

Specialization is a rational response to complexity. You need people who understand customer identity, enterprise architecture, creative workflow, data, or marketing automation at a depth no generalist marketing team could. But deeper expertise in each part of the system doesn’t create the relationships needed to make the whole system work.

Take personalization. It sounds like a marketing capability, but doing it properly can involve customer data, consent, identity, product information, segmentation, decisioning, content, production, activation, and measurement. Some of those capabilities may sit inside marketing. Others may belong to technology, data, product, commerce, legal, sales, or customer service.

The customer sees one brand, and the technology behaves as a single connected environment, with systems exchanging data and triggering activity across platforms, regardless of who owns the budget or which box appears on the org chart.

The organization still has to manage the boundaries behind that experience. The same customer journey can cross multiple teams, reporting lines, and budgets, each with reasonable but different priorities. Those boundaries exist for good reasons.

Security should care about different things from creative. Legal should occasionally make marketers unhappy. Technology has to think about resilience and architecture in ways a campaign team doesn’t, while product and commerce have responsibilities extending well beyond this quarter’s marketing plan.

Adding expertise doesn’t resolve how decisions move across functions. Technology requires these capabilities to operate as a system, but the organization still manages them primarily as specialist functions.

McKinsey’s research into marketing operating models illustrates that gap. In its study of 104 C-suite marketing and growth executives in consumer and retail businesses, only 27% considered their organizations to have mature marketing operating models for advanced use cases.

Just 37% reported successful collaboration within marketing and across functions such as ecommerce, product, and sales, while only 12% identified effective and clear governance as a distinctive feature of their operating model. McKinsey also describes how adding new capabilities to existing structures can reinforce silos, duplicate activity, and add layers of decision-making.

That matters because the capability can exist, the people can be good at their jobs, and the technology can work, yet the outcome can still be constrained by decisions made elsewhere. CreativeOps can streamline production and still be held back by poor product information feeding into it. MOps can redesign the workflow without being able to change the underlying enterprise architecture. A martech team can integrate platforms beautifully while having limited influence over the identity, security, or data standards those integrations depend on.

When a technology problem becomes an organizational one

This doesn’t make every broken stack or slow process a structural problem. Marketing still has plenty of weak implementations, duplicated technology, bad data, and processes that should never have been automated in the first place.

The 2026 CMO Survey found that marketing leaders rated their ability to design the broad architecture of their marketing technology at 4.6 out of seven, compared with 4.9 for selecting the right vendor and 4.8 for choosing the right mix of technologies. The scores suggest we’re slightly better at choosing the parts than designing the machine, and some organizations simply need to get better at operating what they already have.

The structural issue emerges when capable teams repeatedly hit constraints created by the allocation of authority rather than a lack of expertise. Marketing can be accountable for an outcome without sufficient influence over the systems, capabilities, and decisions that determine it.

Look at what sits between marketing intent and enterprise infrastructure.

• Marketing usually owns much of the judgment: the brand, customer proposition, creative direction, commercial priorities, and the outcome it wants to create.
• Technology owns much of the underlying infrastructure: cloud, security, identity, enterprise architecture, core data platforms, and foundational AI infrastructure.

Between those two ends sits a substantial operational layer of workflows, integrations, approval paths, data activation, content supply chains, rights, decisioning, and customer journeys. It’s where marketing intent is translated into something the organization can actually execute.

Every technology wave has added something to this layer: another integration, process, team, system, or dependency. What’s happened less consistently is the deliberate design of the relationships and decision rights across the whole. That’s why ownership alone is a poor answer. The CMO doesn’t need to annex IT, data, product, commerce, and legal. Equally, accountability for marketing outcomes can’t be meaningful if marketing has no influence over the shared capabilities that determine them.

This is also where the industry’s enthusiasm for orchestration runs into a more difficult reality. Customer journey, campaign, content, and workflow orchestration all promise to coordinate activity across a complex environment, but software can only automate decisions the organization has made possible.

A platform can move customer data between systems, but it can’t settle an unresolved question about who controls the customer model. Workflow can automate an approval path, but it can’t decide who should have approval authority. As martech gets better at connecting the work, the unresolved organizational decisions around that work become more visible.

AI removes the human buffer

None of this started with AI, but AI makes it much harder to ignore. For years, people have acted as the unofficial integration layer between functions. A campaign manager messages customer service before a retention offer goes out. Someone in creative knows which product image is approved. An operations lead knows which reviews can safely run in parallel when the deadline moves. Somebody remembers that legal rejected a claim last time, even though the rule doesn’t exist anywhere in the workflow.

Corporate life is full of people navigating ambiguity the organization never formally resolved. They chase approvals, ask data which record is right, call product, negotiate exceptions, and generally keep things moving despite the architecture around them. It’s inefficient and often undocumented, but it also provides a human buffer between systems, teams, and decisions that don’t quite fit together.

AI has much less tolerance for that arrangement. An agent needs access to the right data, permission to use it, rules defining what it may do, thresholds determining when it must escalate, and somebody accountable when it gets the decision wrong. Decisions that people negotiated informally become part of the system design. Permissions encode authority, approval rules encode governance, and escalation paths encode accountability.

The risk is clear. CreativeOps deploys a content agent, MOps introduces a journey agent, sales has an outreach agent, and customer service develops its own service agent. Each may be well designed within its own domain, but each also inherits the objectives, access, incentives, and blind spots of the function that created it. Without a coherent operating model across them, automation can reproduce organizational fragmentation rather than remove it. The silos haven’t disappeared because they now have APIs.

This is why AI changes the urgency of the discussion without changing its underlying cause. Connecting two systems while leaving humans to manage the relationship is one thing. Allowing software to act across them is another. As execution moves into automated workflows, organizations have less room to rely on institutional memory and informal negotiation to bridge gaps in authority, ownership, and process.

The next constraint may sit outside the stack

For CMOs, the challenge extends to growth, customer experience, personalization, commercial performance, and brand consistency — outcomes that depend on capabilities beyond the marketing function. Delivering them requires coordination across the teams that control those capabilities.

That doesn’t imply another restructure, a new empire inside marketing, or a single function taking control of the whole system. Product models, fusion teams, centers of excellence, and other cross-functional approaches already show that organizations can create shared accountability without redrawing every reporting line. The important point is more basic: when an outcome depends on capabilities across several functions, the organization needs decision rights and operating relationships that span the same boundaries.

Martech, MOps, and CreativeOps are particularly important to this discussion because they already operate close to many of these joins. Their value comes from running their respective disciplines well and understanding how those disciplines connect to the wider production system. The same is true of technology, data, product, and commerce. None can optimize the complete customer or commercial outcome in isolation.

As technology moves from connecting systems to acting across them, organizations need to define authority, governance, and accountability across the same boundaries.

The next evolution of marketing may have little to do with adding another technology capability. The stack will continue to evolve, but extracting value from it will depend on whether the organization surrounding it can operate with the same level of connection.

Marketing has outgrown the organization built to run it. The challenge now is whether the operating model can catch up with the production system we’ve already created.