The Next AI Pilot May Be the Wrong Investment

Why It Matters

A pilot has a legitimate purpose when it reduces uncertainty. It can show whether a capability works for a defined task, with selected users and under bounded conditions. Once feasibility is established, the decision changes from “can it work?” to “what should we fund next?” That question requires different evidence.

McKinsey’s 2025 global survey reported widespread AI use but limited enterprise-level EBIT impact from generative AI. It also found workflow redesign had the strongest relationship with reported EBIT impact among the organizational attributes it tested. The survey does not establish causality, but it reinforces why executives should examine the conditions around AI deployment before treating pilot activity as proof of enterprise value.

DigitalQatalyst’s Enterprise AI Value Gap research makes the same evidence boundary explicit: a pilot can establish feasibility under defined conditions, while enterprise value requires evidence that the result can survive normal operating conditions and support a scale, correction or stop decision. The existing portfolio should therefore become part of the evidence base for the next commitment.

The practical implication is simple. A proposed investment should have a clear job: test a genuinely new opportunity, resolve a constraint, strengthen an existing capability, or establish evidence needed for a scale decision. If leadership cannot name that job, the portfolio may be adding activity without improving the investment case.

Make the Portfolio Answer the Funding Question

Before approving the next pilot, use the existing portfolio as evidence. For the material AI investments already underway, ask what outcome was expected to change, what has actually changed, and what remains unresolved. Then ask whether the proposed investment addresses that unresolved issue.

If it does, another pilot may be the right decision. If it does not, redirect the next commitment toward the constraint with the greatest consequence. The Enterprise AI Value Gap uses six conditions to locate where value is being lost; here, its purpose is to identify the constraint that should shape the next investment.

Digital Transformation 2.0 applies the same discipline to investment: funding should follow evidence about outcomes, operating change and unresolved constraints.

The Executive Test

At the next AI portfolio review, put one question on the funding paper: What unresolved value gap in our existing portfolio does this proposed investment address?

Require a clear answer before approval. If the proposed pilot does not resolve a defined uncertainty, constraint or value question, redirect the investment to the evidence gap that does.

Find the gap that should shape your next AI investment →