The shift from implementation cost to discovery cost

With the rapid advancement of AI, implementation costs are decreasing significantly. This shift is creating a growing need for roles that can translate business needs into actionable technology opportunities. I propose defining this role as a Discovery Manager or, alternatively, an Innovation Manager, as an emerging high-value function within modern, large-scale shipping organisations.

The Discovery Manager should be responsible for identifying and clearly articulating business needs by engaging with domain experts across the organisation. This requires strong communication skills combined with a deep understanding of how the business operates in practice. At the same time, the role demands a high level of AI literacy, including familiarity with AI agents, tooling, and rapid prototyping capabilities, in order to map these needs to concrete and feasible technology opportunities.

A central responsibility of the role is to assess and validate value. Against the backdrop of rapidly declining implementation costs, the volume of fast implementations should increase significantly. In the context of AI, the economics of development are fundamentally changing. For many use cases, the primary cost is no longer implementation, but discovery.

Tasks such as competitor analysis, for example, may only require clear problem framing, structured inputs, and access to relevant data sources. Once these elements are in place, an AI-powered solution can often be implemented within hours. This shift makes the ability to define the right problem and scope far more valuable than the ability to build the solution itself.

The Discovery Manager must also ensure that larger initiatives are properly prepared with supporting evidence and handed over with clear specifications and preliminary estimates before entering the standard portfolio intake process and product portfolio management. This ensures that only validated and well-understood initiatives consume organisational capacity.

In addition, the Discovery Manager operates in close alignment with Enterprise Architecture to ensure consistency in tooling and approach, as well as a clear understanding of the scope and structure of enterprise data sources.

AI enables a thin layer of rapid implementations that can be deployed across the organization. With the right enablement, each employee can become significantly more effective in their daily work. Many tasks that previously required dedicated development resources can now be handled directly by business users through AI-assisted workflows, dashboards, and lightweight applications.

An often overlooked emerging advantage is the role of long-tenured employees. Individuals with deep business context, understanding operations, edge cases, and historical decisions, become significantly more valuable when paired with Discovery facilitation. Their ability to frame the right problems, evaluate relevance, and guide solution direction is difficult to replace. By equipping these employees with AI tools and discovery practices, organisations can unlock substantial productivity gains.

This shift has direct implications for development teams. While most organisations already have internal development capacity or rely on external partners, AI enables substantial productivity gains. The role of the developer is evolving from primarily writing code to orchestrating solutions, refining outputs, and handling more complex technical challenges that cannot be easily automated.

As a result, there is a broader transition from traditional backend, frontend, or full-stack roles toward what can be described as full business stack profiles. These individuals combine business understanding with technical capability, allowing them to identify needs, implement solutions, and iterate quickly. They operate across the full lifecycle from problem identification to execution using AI as a core enabler, in close coordination with the Discovery Manager.

Ultimately, the primary constraint is no longer implementation capacity, but the ability to identify the right problems, validate direction quickly, and act with speed and precision. Organisations that succeed in the AI era will be those that invest in discovery as a core capability and systematically increase their learning velocity.

In the AI era, leadership is less about executing large programs and more about identifying the right problems and validating them quickly. As implementation costs fall, the constraint shifts to discovery, understanding what to build and why. This elevates the Discovery Manager into a critical role, enabling organisations to move from idea to evidence at speed.

The leadership challenge is to institutionalize fast, evidence-driven decision-making without losing control. Discovery must operate with enough freedom to explore, while remaining aligned with enterprise architecture to ensure consistency in tooling and data usage. This balance prevents fragmentation and creation of AI slop while preserving speed.

Ultimately, competitive advantage comes from learning faster than competitors. Organisations that embed discovery as a core capability, measured by how quickly they test, validate, and scale ideas, will outperform those still optimised for slow, implementation-heavy models.