U.S. Companies Losing 2.4% of Revenue on Failed AI Projects, Emergn Report Finds
Key Takeaways
- ▸U.S. firms waste 2.4% of annual revenue on failed AI projects due to poor governance and lack of accountability
- ▸Only 30% of organizations treat project shutdowns as normal practice; most continue funding until sunk costs accumulate
- ▸Average organization manages 6+ concurrent transformation initiatives with limited oversight; 1 in 10 have zero formal governance
Summary
A new report from Emergn, a technology and management consultancy, reveals that U.S. organizations lose an average of 2.4% of their annual revenue on AI initiatives that fail to deliver expected value. The survey of 700 senior business leaders shows that poor governance, lack of accountability, and limited visibility into ongoing projects are driving significant waste—with only 30% of organizations considering it normal practice to shut down underperforming initiatives.
The research highlights critical structural gaps in how companies manage AI at scale. The average organization operates more than six transformation and AI initiatives simultaneously, while 1 in 10 operates without any formal oversight structure. Projects are frequently kept alive due to sunk costs and organizational politics rather than tangible evidence of business outcomes, and only 27% of leaders can provide real-time visibility of all programs to their board.
The report argues the solution lies in establishing clear pre-launch criteria, creating accountability structures tied to business outcomes, and normalizing the decision to stop underperforming work early. Experts recommend organizations define three things before funding any initiative: what it's meant to prove, what would indicate success, and what would signal the need to stop—then measure progress by evidence rather than time invested.
- Only 27% of leaders have real-time visibility into all transformation and AI programs; 1 in 5 receive artificially optimistic status reports
- Solution requires pre-launch discipline: defining what to prove, success metrics, and failure thresholds before funding begins
Editorial Opinion
The 2.4% revenue drain reflects a fundamental disconnect between AI adoption pressure and execution discipline. This isn't a technology problem—it's a governance problem rooted in sunk-cost thinking and organizational fear of acknowledging failure. Companies that succeed with AI will be those that establish clear success criteria upfront and make pragmatic stop/go decisions before politics and past investment cloud judgment. The imperative is clear: structure accountability before funding, not after.
