Businesses are spending more time and money on AI, but many are still struggling to turn experiments into dependable, measurable value. That does not mean AI is useless: survey findings point to a gap between trying a tool, embedding it in everyday work and proving that it improves business results.
Why is AI not delivering ROI for businesses?
“AI kind of sucks” captures a real frustration, but it is too broad as a verdict. The more precise problem is that adoption has moved faster than consistent, scaled business value—and many organizations have trouble measuring that value.
In a Q4 2023 survey of 644 participants from organizations in the United States, Germany and the United Kingdom, 49% named difficulty estimating and demonstrating the value of AI projects as a leading adoption obstacle, according to Gartner’s 2024 findings. A tool may save an employee time without reducing costs, increasing revenue or improving a customer outcome. Unless the organization establishes a baseline and tracks the relevant result, local productivity is not proof of company-wide return.
Measurement can also be slow. In a 2025 survey of 100 U.S.-based C-suite and business leaders at organizations with at least $1 billion in annual revenue, none believed their organization had reached the point of measuring GenAI ROI; 31% expected to be able to do so in the following six months, KPMG reported. Those figures describe executives’ expectations and beliefs at the time of that survey, not a universal measure of actual returns.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
#1 Best Overall
As Steve Chase, KPMG’s vice chair of AI and digital innovation, put it in January 2025: “The dynamic nature of AI demands new ways to measure value—beyond the limits of a conventional business case. As leaders work to define the right metrics, those measures must be tightly aligned with the business strategy and should account for the cost of not investing.”
Why do AI pilots fail to make it into production?
A promising demonstration is only an early milestone. A pilot can work with selected examples and close supervision, while a production system must handle routine variation, connect to existing processes and meet security and governance requirements. Even a production deployment may not become a tool employees use consistently or produce a measurable business outcome.
Rank #2
Gartner reported in 2024 that an average 48% of AI projects made it into production, and that the move from prototype to production took eight months. These findings cover AI projects generally, not GenAI alone. The figures are a reminder that technical proof-of-concept success does not automatically translate into an operating system or return.
Adoption studies also capture different stages and populations. In a 2025 study of 150 executives at companies with more than 250 employees across five European countries and several industries, Roland Berger found that 27% of respondents’ companies had fully integrated GenAI into operations and workflows. The study included companies that had already experimented with GenAI, so its results should not be read as a measure of every business.
What are the biggest barriers to using AI in a business?
Organizations face practical constraints beyond model performance. Survey respondents point to data quality, integration, employee readiness and the work of managing privacy and security.
- Data quality and context: In KPMG’s survey of large U.S. organizations, 85% cited organizational data quality as an anticipated challenge, and 71% cited data privacy and cybersecurity. Roland Berger respondents also named data issues as an implementation challenge (28%).
- Integration: 25% of Roland Berger respondents cited integration complexity. A system that does not fit the organization’s existing applications and processes can remain a separate experiment instead of becoming routine work.
- Skills and employee adoption: Roland Berger respondents cited difficulty finding AI or data experts (15%); in KPMG’s survey, 46% cited employee adoption as an anticipated challenge. Training and workflow design matter alongside technical expertise.
- Governance and operating costs: Privacy, cybersecurity, risk controls, monitoring and ongoing maintenance all affect whether a use case is viable. Gartner’s Leinar Ramos said in May 2024: “As organizations scale AI, they need to consider the total cost of ownership of their projects, as well as the wide spectrum of benefits beyond productivity improvement.”
These percentages come from different surveys and should not be ranked as if they measured the same group or conditions. Taken together, they show why an AI initiative can stall even when the underlying technology appears capable.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Is generative AI actually improving productivity at work?
The evidence is mixed, partly because surveys ask different questions and examine deployments at different stages. S&P Global’s 2025 findings indicate greater abandonment: the share of surveyed organizations abandoning a majority of AI initiatives before production rose from 17% to 42% year over year. Respondents also said an average of 46% of projects were scrapped between proof of concept and broad adoption.
In the same S&P Global findings, 46% of respondents whose organizations had invested in GenAI said no single enterprise objective had received a “strong positive impact.” That does not mean those organizations received no benefit at all; it means they did not report a strong positive impact on any one objective.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallBest Value
By contrast, Deloitte’s 2025 survey found that almost all organizations reported measurable ROI for their most advanced scaled GenAI initiatives, while 20% reported ROI of 31% or more. This is self-reported ROI for respondents’ most advanced scaled initiatives—not a result for all pilots, all adopters or every use of GenAI.
The apparent conflict makes more sense when the comparison is kept specific: a project dropped before broad adoption is not equivalent to an advanced, scaled initiative. Deloitte’s Joe Ucuzoglu, global CEO, described the shift in its 2025 State of Generative AI Q4 release: “GenAI use cases are rapidly proliferating in leading enterprises across industries. We are seeing a shift as leaders move past the initial hype to strategically deploying GenAI in the core of their businesses. Focus is essential, prioritizing demonstrated use cases with measurable return on investment.”
What can businesses learn from the gap between experimentation and value?
Gartner describes AI-mature organizations as tending to invest in operating models, AI engineering, upskilling and change management, and trust, risk and security capabilities. These are reported differences, not a guaranteed recipe for returns. The practical lesson is to assess a use case as an operating change—not just a model or software purchase.
Before piloting or expanding a business AI system, leaders can ask:
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →- What specific business outcome should improve: time, cost, revenue, customer experience or risk?
- What is the baseline, and how will the organization measure change against it?
- Does the system fit the real workflow and the data employees need to use it?
- Who checks output quality, privacy, security and compliance—and what happens when the system is wrong?
- What are the full implementation and ongoing operating costs, including integration, oversight and training?
- What evidence would justify scaling, changing or stopping the use case?
Roland Berger’s Edeltraud Leibrock, global managing director, emphasized the data foundation in May 2025: “The full potential of AI can only be unlocked by bringing together structured and unstructured data in context across enterprise processes,”
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
