Why AI Adoption Doesn't Equal ROI: The Value Realization Problem
- Nancy Wang
- 2 days ago
- 6 min read
Updated: 7 hours ago
AI adoption has gone nearly universal: 88% of organizations now use it in at least one business function. Yet only 39% can attribute any measurable impact to enterprise-level EBIT, and most of those report an impact below 5%, per McKinsey's State of AI 2025. Only about 6% of organizations qualify as what McKinsey calls "AI high performers."
The MIT NANDA report, The GenAI Divide: State of AI in Business 2025, went further, concluding that 95% of enterprise generative AI pilots fail to deliver measurable ROI. BCG's research tells a similar story: 60% of companies generate no material value from their AI investments, while only a small minority captures substantial value at scale.
The spending behind these results is not just large. It is accelerating. Global enterprise spending on AI solutions will more than double from $307 billion in 2025 to $632 billion by 2028, per IDC forecasts.
Across years of consulting at Booz Allen and later with media and enterprise clients, I've seen this pattern repeat with every new technology wave: investment and activity surge ahead of the strategic clarity required to convert them into outcomes. This does not mean AI is failing. It means many organizations are still learning how to identify, prioritize and measure the initiatives most likely to create value.
The challenge is not adoption. The challenge is value realization.
The Problem of Too Many Opportunities
Most organizations are not suffering from a shortage of AI ideas. If anything, they have the opposite problem.
Every department can identify dozens of use cases. Marketing wants content generation. Customer support wants AI-powered service. Operations wants automation. Product teams want intelligent features. Leadership wants productivity gains.
Individually, many of these ideas appear promising. Collectively, they create a prioritization challenge that few organizations are equipped to navigate. Resources, budgets and talent are all finite. No organization can pursue every opportunity simultaneously, yet many continue to evaluate AI initiatives one at a time rather than as a coherent portfolio of investments.
The consequences are predictable. Projects move forward based on enthusiasm, executive sponsorship or the perceived urgency of catching up to competitors rather than a disciplined assessment of expected value. The MIT NANDA researchers observed that companies often misallocate AI investment toward highly visible front-office initiatives in sales and marketing, even though back-office automation tends to deliver more reliable ROI.
The result is a pattern that should concern any leader: visible projects accumulate budget while less glamorous but higher-value opportunities go underfunded. Only 15% of AI decision-makers report a positive impact on profitability over the past year, Forrester finds, and the firm predicts companies will defer roughly 25% of planned 2026 AI spend into 2027 as the gap prompts a market correction.
The organizations seeing the strongest results from AI are not necessarily deploying more AI. They are becoming more deliberate about where they invest.
From ROI to ROAI: Rethinking Returns
Part of the challenge stems from the way many organizations still apply traditional ROI thinking to AI investments. A more useful frame is ROAI, or Return on AI: a measurement model that accounts for the multidimensional value AI actually creates.
Historically, technology investments were justified through relatively straightforward calculations. A new system reduced costs, eliminated manual effort or generated incremental revenue. The return could be modeled and measured with reasonable confidence, and the business case lived or died on a single number.
AI is more complex. While some initiatives produce direct financial returns, others create value indirectly by increasing organizational capacity, accelerating decision-making, improving customer experiences or enabling capabilities that did not previously exist.
Leaders who evaluate AI through a one-dimensional lens often miss the most important sources of value, and they sometimes fund the wrong projects because the wrong projects are the easiest to model.
A more useful approach considers value across four distinct dimensions.
Revenue Growth is the most familiar dimension. Can AI help generate more pipeline, increase conversion rates, improve retention, accelerate product development or enable expansion into new markets? McKinsey's research shows revenue gains from AI concentrate in marketing and sales, in strategy and corporate finance and in product development.
These outcomes tend to be the most visible because they show up at the top of the income statement. They are also the hardest to attribute cleanly, since revenue is rarely the result of any single intervention.
Operational Efficiency is where most organizations start, because the benefits are easier to identify and measure. AI can reduce manual effort, streamline workflows, shorten cycle times and increase employee productivity. Organizations are seeing 10% to 20% cost reductions in software engineering and IT, with similar gains emerging in manufacturing operations, McKinsey reports.
Efficiency initiatives may not immediately generate new revenue. They can significantly improve operating leverage and free skilled teams to focus on higher-value work.
Effectiveness is frequently overlooked because the benefits resist easy quantification. Better decisions, improved customer experiences, higher-quality outputs, faster insights and greater consistency may not appear on a financial statement in any given quarter, but they often compound into meaningful competitive advantages over time.
The same McKinsey survey finds 64% of organizations report AI is enabling innovation within their business, an outcome that is real and valuable even when it cannot be precisely quantified.
Strategic ROI is the value created by building capabilities the organization will need in the future. Some AI initiatives are less about immediate returns and more about developing the data foundations, operating models, governance structures and organizational muscle required to compete in an AI-enabled market. These investments often look expensive on a spreadsheet and inexpensive on a strategic horizon.
Together, these four dimensions form the foundation of ROAI. Viewed through this lens, the question shifts from whether a project will save money to how an initiative creates value, how confident the organization is in that value and how important it is to the broader strategy.
Why Measuring AI Success Is Difficult
Unlike traditional technology projects, AI initiatives rarely fit neatly into a single category. A content marketing platform may reduce production costs while simultaneously increasing content velocity, improving quality and expanding market reach. An AI assistant may improve employee productivity while increasing customer satisfaction. An intelligent workflow may reduce costs today while creating strategic advantages tomorrow.
This multidimensional nature is precisely why measuring AI success is so challenging. The temptation, when faced with hard-to-measure value, is to default to whatever is easiest to count.
Organizations focus on the metrics that are simple to capture rather than the metrics that matter most. Pilot success becomes "did it launch?" rather than "did it move the business?" Productivity becomes "hours saved on a task" rather than "did those hours produce more valuable output?"
The result is that valuable initiatives may be undervalued while highly visible initiatives receive disproportionate investment. The organization quietly drifts toward an AI portfolio that looks active but is not actually moving the numbers that matter.
From Experimentation to Prioritization
As AI matures, leading organizations are beginning to shift their mindset. The conversation is moving beyond experimentation and toward prioritization.
Instead of asking, "Where can we use AI?" they are asking, "Where will AI create the most value?"
That distinction is subtle but important. The first question generates ideas. The second creates strategy.
Organizations that successfully answer the second question are better positioned to align investments with business objectives, focus resources on high-value opportunities and sequence near-term wins alongside long-term transformation. The organizations capturing the most value from AI are three times more likely to use it for transformative change rather than incremental improvement, McKinsey's research underscores, and they pair efficiency goals with explicit growth and innovation objectives.
A Framework for Better Decisions
At Agentic Foundry, we believe the future of AI success will belong to organizations that treat AI investments with the same rigor applied to any strategic portfolio of initiatives. I've seen the same pattern repeat across very different industries: the organizations that capture the most value from a new technology are not the ones pursuing the most use cases, but the ones making the sharpest choices about what to pursue first.
Not every AI opportunity deserves the same level of investment. Not every use case should be pursued immediately. And not every promising idea will create meaningful business value.
The organizations that realize the greatest returns from AI will be those that develop a structured approach to evaluating opportunities, measuring value across multiple dimensions and prioritizing initiatives based on both business impact and organizational readiness. That requires moving from a use-case mindset to a portfolio mindset and from a single-metric ROI calculation to the multidimensional view of value that ROAI is designed to capture.
Because ultimately, AI success is not determined by how many projects an organization launches. It is determined by how effectively it converts AI investments into measurable business outcomes.
One question to ask: if you ranked your current AI initiatives by expected value across all four ROAI dimensions rather than by visibility or sponsorship, would the funding order hold?
Agentic Foundry: AI For Real-World Results
Learn how agentic AI boosts productivity, speeds decisions and drives growth
— while always keeping you in the loop.
