TL;DR: Nearly a third of enterprises see almost total failure of AI proof-of-concept projects, with poor preparation—not technology—identified as the primary cause. Only 9% of companies successfully move more than half their AI pilots into production.
Omdia’s 2025 AI Market Maturity Survey has revealed a stark divide in enterprise AI adoption: whilst 46% of firms successfully deploy over 10% of their AI projects, 31% experience success rates below 5%. The research underscores that AI implementation challenges stem from inadequate planning rather than technological limitations.
The Planning Gap
The survey identifies poor preparation as the chief culprit behind project failures. Successful AI deployments require organisations to identify specific business challenges before initiating proof-of-concept work, with critical effort needed both before and after pilot testing.
Supporting research from Cisco found that only 32% of companies properly identify which human tasks AI should supplement or replace. US Environmental Protection Agency CIO Carter Farmer noted that organisations are “hastily rushing to deploy AI projects without defining a clear use case.”
Resource Divide
Company size significantly impacts AI success rates. Firms with revenue below $100 million typically run fewer than five proof-of-concept projects, whilst the largest enterprises manage over 100 simultaneously. This resource gap appears to influence deployment success, suggesting AI implementation requires substantial organisational capacity.
Mixed Returns
Despite implementation challenges, 30% of respondents reported AI deployments aimed at productivity exceeded expectations, whilst 49% met expectations. However, proving return on investment remains difficult—earlier Lenovo research found business leaders unconvinced of AI’s value, with demonstrating ROI identified as one of the biggest barriers to adoption.
The survey suggests AI is neither the transformative force some claim nor a complete failure, but rather a complex undertaking requiring careful planning and substantial resources to deliver meaningful results.
Source: The Register