AI Consulting in Practice – NLW, Super ai

Enterprise AI Adoption & ROI: A Deep Dive into Current Findings

Key Concepts:

• AI Bubble Narrative: Current skepticism surrounding the inflated expectations and potential overvaluation of AI technologies.
• Enterprise AI Adoption: The increasing, but uneven, integration of AI technologies within organizations.
• AI Agents: Autonomous entities powered by AI, designed to perform specific tasks within an enterprise.
• ROI (Return on Investment): Measuring the value generated from AI investments, a challenging metric in the current landscape.
• High Performers (AI Adoption): Organizations demonstrating systematic and comprehensive AI strategies, exceeding typical adoption rates.
• Use Case: A specific application of AI technology to address a particular business problem or opportunity.

I. Current State of Enterprise AI Adoption

The enterprise AI landscape is experiencing growth, with nearly all organizations utilizing AI to some degree. A significant inflection point this year has been the surge in AI adoption within coding and software engineering, extending beyond dedicated software teams to other departments seeking to leverage code generation and manipulation. While initial expectations of widespread automation via AI agents haven’t fully materialized, agent adoption is demonstrably increasing.

KPMG’s quarterly pulse survey reveals a jump in full production agent deployment from 11% in Q1 to 42% in Q3 among companies with over $1 billion in revenue. This suggests a faster-than-anticipated transition from pilot programs to operational implementation. This adoption is driving a shift in focus towards human-agent interaction, requiring upskilling and enablement initiatives, and a decrease in initial resistance.

However, a substantial portion of enterprises (62% according to McKinsey’s State of AI study) remain in the experimentation or pilot phase, struggling to scale initial successes. Larger organizations are generally leading the scaling efforts, contrary to expectations that smaller, more agile companies would be quicker to adopt. Adoption patterns are uneven across departments, with IT operations currently leading the way.

II. The Bifurcation of AI Performance: Leaders vs. Laggers

A clear distinction is emerging between leading and lagging organizations in AI adoption. High-performing companies are characterized by a more comprehensive and systematic approach, pursuing multiple AI initiatives simultaneously. They are not solely focused on immediate productivity gains but also explore revenue growth, new capabilities, and product development.

Spend on AI is projected to continue increasing, with KPMG’s pulse survey indicating an expected expenditure of $130 million in the next 12 months (up from $114 million previously). Deloitte data shows over 90% of organizations plan to increase AI spending. This increased investment is accompanied by a growing emphasis on demonstrating tangible impact and ROI.

III. Shifting Expectations Around ROI Realization

Despite media concerns, optimism regarding AI ROI is increasing. A KPMG CEO survey revealed a significant shift in expectations: in 2024, 67% of CEOs anticipated realizing ROI within 1-3 years, compared to 63% expecting 3-5 years in the previous year. Furthermore, 19% now expect ROI within 6 months to 1 year, a substantial increase from prior expectations.

However, accurately measuring AI ROI remains a significant challenge. 78% of survey respondents acknowledge the inadequacy of traditional impact metrics for evaluating AI investments. This prompted a new study to gather self-reported ROI data directly from users.

IV. Initial Findings from the ROI Study (Over 2500 Use Cases)

A study conducted with over 1,000 organizations and 3,500 use cases reveals that organizations are finding ROI in AI, with 44.3% reporting modest ROI and 37.6% reporting high ROI. Only 5% reported negative ROI, though this doesn’t necessarily indicate failure, but rather that costs have exceeded perceived gains. A striking 67% anticipate increased and high growth in ROI over the next year, even among those currently experiencing negative ROI (53% expect high growth).

Key ROI Categories & Findings:

• Time Savings: The most common ROI category (35% of use cases), with a typical savings of 1-10 hours per week (often around 5 hours). Saving 5-10 hours weekly equates to regaining 7-10 work weeks annually.
• Increased Output & Quality Improvement: Follow time savings as dominant ROI categories.
• Organization Size & ROI Focus: Organizations with 200-1,000 employees prioritize increasing output, potentially reflecting their stage of growth.
• Leadership Perspective: Suite-level executives and leaders are more focused on higher-level ROI categories like increased output and new capabilities, and report higher levels of transformational impact. 17% of their submitted use cases already demonstrate transformational ROI.
• Small Organization Impact: Organizations with 1-50 employees are experiencing a disproportionately high level of transformational impact.
• Coding & Software Use Cases: AI applications in coding and software development consistently demonstrate higher ROI and lower negative ROI.
• Risk Reduction – Unexpectedly High Impact: While representing only 3.4% of use cases, risk reduction applications have the highest likelihood of delivering transformational ROI (25%). This is attributed to AI’s ability to handle large volumes of data in compliance and risk management functions.
• Industry Specifics: Healthcare and manufacturing demonstrate higher average ROI compared to other industries.

V. The Role of Agents & Systematic AI Implementation

Automation and agentic use cases are increasingly contributing to ROI, significantly outperforming other use cases. Organizations adopting a systematic, cross-organizational approach to AI implementation are experiencing greater success. The number of use cases submitted by an organization directly correlates with the level of ROI achieved, suggesting that broader AI integration yields greater benefits.

Notable Quotes:

• “The ways that we have measured impact of previous technologies and just previous initiatives are kind of falling flat with AI.” – Speaker, highlighting the challenge of applying traditional ROI metrics to AI.
• “You are seeing a pretty meaningful uptake of of agents inside the enterprise.” – Speaker, emphasizing the rapid adoption of AI agents.

Conclusion:

Despite ongoing concerns about an “AI bubble,” organizations are actively adopting AI and realizing tangible value. While many remain in the pilot phase, a growing number are deploying AI agents and experiencing significant ROI, particularly in areas like time savings, output increase, and risk reduction. A systematic, cross-organizational approach to AI implementation, coupled with a focus on higher-level strategic goals, appears to be a key differentiator between high-performing and lagging organizations. The initial findings from the ROI study suggest a positive outlook for AI investment, with increasing optimism and a growing emphasis on demonstrating measurable impact. Further research is needed to refine ROI measurement methodologies and identify best practices for maximizing the value of AI across diverse industries and organizational sizes.