By Reynand Wu 
By PYMNTS
June 26, 2026
As the initial hype surrounding generative artificial intelligence (AI) gives way to the harsh realities of fiscal accountability, a stark "ROI paradox" has emerged across the global enterprise landscape. While billions of dollars continue to flow into AI infrastructure, large-scale deployments, and pilot programs, a significant number of organizations are finding themselves unable to answer the most fundamental question posed by their boards and CFOs: "Is this actually working?"
Recent insights from the Wedbush Securities Disruptive Technology Conference have brought this issue to the forefront, revealing that many enterprises are scaling AI tools without the necessary frameworks to quantify their impact. This lack of measurement is not merely a technical oversight; it is becoming a significant bottleneck that threatens to stifle the next wave of technological transformation.
The Measurement Vacuum: Core Findings from Wedbush
During the Disruptive Technology Conference held earlier this week, analysts from Wedbush Securities, led by the renowned Dan Ives, delivered a sobering assessment of the current AI investment climate. Their primary finding was clear: the enterprise sector is currently navigating a "measurement vacuum."
According to the investor note released on Friday, June 26, 2026, many organizations have launched AI pilots with high ambitions but little in the way of KPIs (Key Performance Indicators) or benchmarking standards. The result is a fragmented landscape where businesses are deploying sophisticated, high-cost tools without a clear mechanism to distinguish between effective, value-generating implementations and those that are merely consuming resources.
The absence of these frameworks makes it increasingly difficult for management teams to justify ongoing spending. Without empirical evidence of success, identifying which AI approaches are yielding competitive advantages becomes a matter of guesswork rather than data-driven strategy. This uncertainty, in turn, hinders organizational confidence—the very fuel required to pivot from experimentation to full-scale enterprise integration.
Chronology of the AI Investment Shift
To understand the current state of enterprise AI, one must look at the timeline of the adoption curve over the last several years:
- 2023: The Year of Experimentation: Following the public breakthrough of generative AI, companies rushed to integrate Large Language Models (LLMs) into their workflows. The focus was on "speed to market" rather than return on investment.
- 2024: The Infrastructure Build-Out: Investments shifted toward cloud capacity, specialized hardware (GPUs), and data cleaning. During this phase, ROI was often overlooked in favor of achieving technical feasibility.
- 2025: The Reality Check: As budgets for these massive build-outs ballooned, CFOs began demanding greater financial visibility. Research from PYMNTS Intelligence began to highlight that executives were starting to reconcile their expectations with the long-term nature of AI maturity.
- 2026: The Accountability Pivot: The current moment marks a transition toward "value-based AI." The conversation has moved from "What can AI do?" to "How does this contribute to the bottom line?" The Wedbush findings underscore that many firms are currently ill-equipped for this necessary transition.
Supporting Data: The Disconnect Between Expectation and Reality
The struggle to measure ROI is compounded by the fact that the benefits of AI are often non-linear and long-term. PYMNTS Intelligence research, as noted by CEO Karen Webster in September 2025, found that the majority of enterprise executives have adopted a pragmatic, long-term outlook.
When surveyed on their expectations for positive payback from generative AI investments, over 80% of executives indicated that they anticipate a timeframe between three and ten years. This suggests that the frustration from boards may be rooted in a misalignment of timelines: the board seeks quarterly results, while the AI transformation is a multi-year, foundational shift.
Organizational Barriers to Performance
The "Enterprise AI Readiness Gap" report from PYMNTS Intelligence further clarifies why measurement—and by extension, success—remains elusive. When asked what constrains AI performance more—technological capability or organizational readiness—a resounding 71% of executives pointed to the human and structural side of their businesses.
The report identifies several persistent bottlenecks:
- Data Quality: AI models are only as good as the data they ingest. Disparate, siloed, and "dirty" data remain the primary technical hurdle.
- Budgetary Constraints: The high cost of compute and talent is clashing with the slow, uncertain payback periods.
- Governance Processes: Rigid corporate structures are struggling to adapt to the agile, iterative nature of AI development.
- Talent Gaps: The lack of internal expertise to manage AI lifecycles exacerbates the difficulty of creating sound measurement models.
Official Responses and Executive Sentiment
The pressure is mounting. Dan Ives noted in the Friday investor note that the feedback from the front lines of the tech industry is becoming increasingly urgent.
"Many executives noted that customers are feeling increased pressure from their boards and CFOs to demonstrate actual returns from AI," Ives stated. "The inability to answer this question presents a real barrier to additional investments in long-term technological buildouts."
Karen Webster’s analysis echoes this sentiment, framing the challenge as a fundamental misunderstanding of the nature of digital transformation. "These enterprise executives also understand that big-‘T’ transformation doesn’t usually happen on a predictable timetable, nor with the expectation of an immediate or direct payback ‘in the millions,’" Webster wrote.
The core issue, therefore, is not necessarily that AI is failing, but that the enterprise is failing to integrate AI into its existing performance management systems. Companies are attempting to measure a new, revolutionary technology using the same KPIs they applied to legacy IT upgrades, which leads to a distorted view of value.
Implications: The Path Forward
The findings from Wedbush and PYMNTS Intelligence suggest that we are at a critical juncture. The "piecemeal" approach to problem-solving—fixing one silo or adding one feature at a time—will no longer suffice. To survive the current cycle of skepticism, enterprises must move toward a holistic operating model.
1. Synchronized Transformation
The report emphasizes that improvement cannot be siloed. "Improve data quality, clarify responsibility, address talent gaps and rethink budgets in parallel to take full advantage of cross-functional AI operating models," the report advises.
2. Defining "Success" Beyond the Dollar
Enterprises must move toward a more nuanced definition of ROI. In the early stages of AI adoption, metrics such as "time saved on repetitive tasks," "improvement in customer service response accuracy," and "speed of document synthesis" are more indicative of long-term success than immediate revenue spikes. By failing to quantify these operational efficiencies, companies are ignoring the "hidden" returns that will eventually fuel profitability.
3. The Need for Governance
The most successful organizations are now implementing robust AI governance boards that report directly to the C-suite. These boards are tasked with setting benchmarks, ensuring data integrity, and bridging the communication gap between technical teams and financial stakeholders.
Conclusion: The Maturity Phase
The current anxiety surrounding AI ROI is a sign of a maturing market. The "gold rush" phase is over; the "operationalization" phase has begun. While the lack of current benchmarks is a significant hurdle, it is also a catalyst for necessary change.
As enterprises begin to treat AI not as an external "add-on" but as a core component of their business infrastructure, the methods for measuring its value will evolve. Those who successfully implement frameworks to track both the operational efficiency and the strategic value of their AI investments will likely emerge as the market leaders of the next decade. Conversely, those who continue to invest without a clear roadmap—or a way to measure the distance traveled—risk being left behind in a landscape that no longer tolerates the "black box" approach to enterprise spending.
Lina Hope 
Layla Zulfa 
Siti Muinah 