By: Jeff Mullen, VP of Strategy at Prodoscore
Boards are approving millions of dollars in AI investments without a reliable way to determine whether those investments are improving business performance. Most organizations can tell you who is using AI, how often they log in, and how many prompts they generate. Far fewer can explain whether AI is helping employees produce better work, complete it faster, or reduce costs.
That disconnect reflects one of the biggest blind spots in enterprise AI today. Companies have become skilled at measuring adoption and remain surprisingly poor at measuring output. As AI moves from experimentation to a meaningful operating expense, that distinction is no longer an IT or HR issue but a board-level one.
The scale of that blind spot is easy to underestimate. People costs are the largest line item on most companies’ income statements, with the U.S. Bureau of Labor Statistics reporting that wages and salaries alone account for roughly 70 percent of total employer compensation costs. Any function responsible for that much spend eventually answers to finance, which is exactly why AI has changed what people do at work and who is accountable for measuring it. A growing share of that work is now executed end-to-end by agents, with no human in the loop. Once autonomous systems produce outcomes and incur costs on their own, the conversation can no longer stay with HR.
For most companies, AI adoption has outpaced the ability to measure its impact. Spending on AI tools and agents has climbed across nearly every industry, and the instinct has been to treat wider access as inherently good.
Adoption is easy to measure, since a dashboard can show who logged in, how many queries were run, and how many tokens were used. Output is harder to pin down because it requires knowing what the work actually was and whether the result was better, faster, or cheaper than work without AI. The gap between those two numbers is where AI’s business case either holds up or falls apart.
The blind spot in unlimited access
Giving employees broad access to AI made sense when organizations were encouraging experimentation. It becomes a much riskier strategy once AI spending becomes a meaningful operating expense.
Unlike traditional enterprise software, which has predictable licensing costs, many AI platforms charge based on usage. Two employees doing similar work can generate vastly different costs depending on how they use the technology. Without visibility into that variation, there is no way to distinguish efficient use from unnecessary expense.
Giving every employee unlimited, unmeasured access is not a benefit; it is an open-ended financial commitment with no clear way to connect spending back to business outcomes. Organizations need to become more deliberate about which tools employees use, how they are used and where they create measurable value.
Why high adoption doesn’t mean high output
One of the most expensive assumptions organizations can make is believing that high AI adoption automatically translates into better business performance.
An employee can run dozens of AI-assisted tasks in a day without producing more value, particularly if the underlying workflow was never redesigned around the tool, or if the tasks being automated were not the ones limiting output in the first place. High usage can just as easily reflect inefficient habits or low-value work being run through an expensive tool simply because it is available.
AI is delivering real value in many functions, but anecdotal evidence of productivity gains is not a measurement framework.
Boards making investment decisions need evidence that connects AI-enabled work to financial outcomes, such as cycle time, error rate, revenue per employee, and cost per unit of output. Without those connections, AI investments are evaluated on assumptions rather than evidence, a practice that becomes increasingly difficult to justify as scrutiny over technology investment grows.
Building the measurement framework before the pressure hits
The companies that navigate the AI transition well are getting ahead of this question rather than waiting for it to surface in a budget review. A few practices are worth putting in place now.
Separating adoption metrics from output metrics and reporting them distinctly is the most important starting point. Usage data measures reach, not value, and the two need to be tracked independently. Equally important is establishing a baseline before scaling access further, because without knowing what a workflow’s output looked like before AI was introduced, there is no credible way to measure what changed after.
Measurement should also be tied to financial outcomes the organization already tracks, rather than new metrics invented for the occasion. Metrics like cost per output, revenue per employee, and cycle time are already part of how boards evaluate performance, and AI measurement should connect to those same standards.
Workforce and workflow data provide the foundation for answering those questions. This is not a new discipline, as organizations have long used people analytics to connect labor spend to business outcomes, but AI has raised the stakes considerably. Organizations that begin building that capability now will be able to justify AI investments with evidence, not estimates.
AI adoption is becoming table stakes, and measuring its business impact is the differentiator that will separate disciplined operators from the rest.
About the Author
Jeff Mullen is VP of Strategy at Prodoscore, a workforce productivity analytics platform. He has spent his career at the intersection of finance and people analytics, building measurement functions at organizations including the Federal Reserve Board and LinkedIn and advising enterprise leaders on how workforce data connects to business outcomes. His work focuses on helping organizations build the visibility they need to make better decisions about people and performance.
About Boardroom Pulse
Boardroom Pulse is the C-suite’s trusted source for forward-thinking, insightful coverage on corporate governance and the latest developments shaping today’s business world.
Our mission is simple yet ambitious: elevate governance standards and empower modern business leaders. To achieve this, we deliver comprehensive, timely news, in-depth analysis, and thought leadership that spark dialogue, highlight best practices, and promote responsible leadership in boardrooms and executive suites nationwide.
