In my previous article, I argued that Model Context Protocol (MCP) represents a meaningful structural shift in sales enablement. The central idea was not simply that AI would improve access to knowledge, but that it would begin to close the gap between knowledge and execution by enabling systems to engage more directly in workflows. The implication was significant: enablement was moving from providing content to supporting action.
However, once that connection is achieved, a more fundamental issue makes itself known. Access, for all its importance, was never the root constraint. It was simply the most visible one. MCP removes much of that friction. What it does not resolve is how knowledge itself has been constructed.
Over the past decade, sales enablement has accumulated a vast estate of content. Methodologies, product narratives, customer stories, pricing guidance, and coaching frameworks sit across a range of systems and formats. As noted previously, the challenge has never been the absence of knowledge; most organisations have more than they can realistically use. The difficulty lies in converting that knowledge into something that can be applied precisely when it matters.
That difficulty persists even in an AI-enabled environment, because most of what we have built has been designed for human consumption rather than machine utilisation. Courses, decks, documents, and videos are effective for communication, but they are inherently static. They assume a user will interpret, translate, and adapt the content to a given situation. That expectation is manageable in a training environment. It becomes far more problematic in the flow of work.
When an AI agent retrieves a piece of knowledge in its current form, it does not receive a clearly defined set of instructions or a framework that can be directly applied. Instead, it encounters material that must be interpreted, reconstructed, and in some cases inferred. At small scale this may appear sufficient. At enterprise scale, it introduces inconsistency, reduces precision, and ultimately constrains the reliability of the system.
In other words, MCP solves the problem of connection, but it does not solve the problem of representation.
This is where the next evolution in sales enablement begins.
Want to read the full article? Click here