The Model Context Protocol promises a more manageable ad market. Instead of full unification, its real value lies in systematizing granular data and automating repetitive tasks. Here’s what that means for media operations.
Media and publishing professionals looking for a single, unified ad market may be disappointed by the Model Context Protocol (MCP). According to Magnus Ohlin, co-founder and Chief Innovation Officer at Camphouse, MCP is unlikely to deliver true market standardization. Instead, its strength lies in making complex media environments more manageable by enabling systematic, granular data use and automating repetitive operational tasks.
Ohlin notes that MCP does not signal a new era of market unification. Multiple protocols are emerging at once, and each market participant continues to build their own logic, environment, and setup. As Paul Ripart of Prisma Media has pointed out, this fragmentation could even create new, less visible walled gardens. The idea of running seamless, cross-category campaigns remains out of reach due to differing buying logics, technical restrictions, and platform rules. MCP does not resolve these core divides.
The protocol’s real potential, Ohlin argues, is in automating micro-tasks at scale. These include categorizing actions, applying consistent taxonomies, generating clean UTM parameters, structuring campaigns with clear naming conventions, and ensuring the right information reaches the right platform. While these tasks are simple, they are repetitive and still largely manual. MCP enables automation and consistency, turning scattered micro-tasks into structured, actionable data.
Today’s media landscape is rich but tangled. All the necessary data exists, but it is often knotted and hard to interpret. To unlock value, organizations must be willing to untangle this complexity and regain granular control-understanding each campaign, metric, and interaction individually. MCP becomes relevant at this level, not by unifying everything, but by activating each data thread in the right sequence and scale. Once this groundwork is done, a readable and actionable network of data can emerge.
Another key shift is moving beyond traditional API thinking. While MCP can tightly connect the media ecosystem, agent-based approaches go further. For example, the combination of Claude Cowork and Claude for Chrome demonstrates how agents can challenge established workflows by interacting directly with APIs and browser interfaces. Many current processes rely on limited integrations within separate SaaS tools and browser tabs. Agentic AI, from MCP to virtual actions on interfaces, enables direct, universal interaction with nearly any platform.
This expanded agent spectrum shifts the focus from technical logic to user experience, fundamentally changing how campaigns are managed. However, this only works if the underlying data is clean. MCP does not correct errors-it simply executes. Without disciplined data hygiene, automation can amplify mistakes. Clear rules, preconfigured accounts, and defined buying frameworks are essential to ensure consistency and prevent automation from becoming counterproductive.
MCP’s strength is interoperability, but not in the way many expect. Rather than creating a single, all-encompassing system, MCP allows agents to coordinate defined actions across multiple tools and platforms. This could mean connecting media planning tools like Asana or Monday.com, automatically generating media plans from briefings, or distributing data to the right platforms. Interoperability happens where people already work, respecting existing workflows instead of forcing standardization. For example, a media plan can be created within Claude and executed in a media operations platform, or agents can work in the background to enhance familiar processes.
Ultimately, the media market is not becoming simpler, but it can be better organized. MCP and agentic AI-including browser-based agents-can help untangle operational complexity, structure actions, optimize media buying at the task level, and make data truly actionable. The key is accepting that complexity cannot be solved by unification alone; only through structure does it become useful. As seen in other experiments with editorial workflow innovation, such as Publish X’s approach to scalable editorial judgment, the focus is shifting from connecting more systems to organizing what already exists-avoiding the risk of simply automating chaos.
