A recent MIT report confirms a stark reality. U.S. businesses have invested up to $40 billion in artificial intelligence, with 95% achieving almost no measurable return. This is not a rounding error on an IT budget. It is a strategic miscalculation at the highest levels of leadership.
The failure is not in the technology, but in its application. Many organisations are practising a modern “cargo cult”, performing the rituals of innovation without understanding the principles of value. The MIT study proves it. Projects are stalling because they are disconnected from the operational reality of the business.
This is the expensive consequence of confusing activity with progress.
The Anatomy of Failure: Theory Over Results
The study shows that failed initiatives are often aimed at the wrong targets and are not integrated into actual workflows. This mirrors what we see consistently. A solution is built in an innovation lab but never survives contact with the core business. It exists as a theoretical project, detached from the teams expected to use it and the commercial outcomes it is meant to drive.
This academic approach is a well-worn path to failure. It treats transformation as a subject to be studied in committees, rather than a result to be delivered on the ground. It prioritises the elegance of the plan over the effectiveness of its execution.
The Mandate for Success: Delivery Over Deliberation
The 5% of companies succeeding with AI are not operating differently. They are built differently. Their approach is not academic, it is commercial. The report finds they “pick one pain point and execute well”.
This is the signature of an organisation that values outcomes over overhead. For example, we have seen this principle deliver dramatic results in highly regulated sectors. A financial services firm, which had struggled for a year with a complex data automation project, changed tack. They mandated a single, senior practitioner to solve one specific regulatory reporting issue. The result was delivered in eight weeks.
This is what happens when you prioritise delivery over deliberation. Success is not found in owning a tool, but in the skill of its application. It is this operational grip that turns technological potential into commercial reality.
The MIT report is more than a warning. It is a clear instruction. The market is saturated with advisors who can talk about AI strategy. The evidence shows that what creates value is the rare ability to connect that strategy to a tangible, commercial outcome. The question is no longer ‘What can AI do for us?’.
It is, ‘Who has the operational grip to ensure it delivers?’
The COGENT Principle
Insight without execution is an academic exercise. At COGENT, our work is built on a single premise: technology does not deliver results, disciplined leadership does.
We don’t sell technology or theoretical strategy. Our entire model is focused on providing the operational grip required to turn a complex business ambition into a measurable commercial outcome. We deploy senior practitioners, not junior analysts, to ensure the work gets done.
To bridge the gap from insight to action, we have two established starting points for leadership teams:
1. Enterprise AI Readiness Field Guide. This is a three-week programme designed to produce a board-grade 90-day plan. It includes audits, workshops, and roadmaps that deliver your top five use cases, clear owners, and governance you can defend. The first step is a 45-minute executive scoping call, during which we work with you to develop an initial list of practical use cases that reflect your business’s and your sector’s current realities.
2. AI in a Day: Hands-On Lab. This focused workshop ensures your leadership team is working from the same map. Your team will demystify the technology, experiment with safe tools, and leave with 2–3 owned use cases and a clear 30/60/90 plan. Within five days, you receive the summary pack and a starter safe-use policy to maintain momentum.