McKinsey Study Shows Scaling AI Adoption Amid ROI Hurdles and Rising Infrastructure Costs

Enterprise adoption of agentic artificial intelligence is accelerating across major corporations even as organizations encounter ongoing friction in translating these deployments into clear financial returns, according to recent survey data from global management consulting firm McKinsey first reported by TechRadar Pro. Despite the hurdles surrounding return on investment, the study indicates that businesses are beginning to see measurable operational benefits and cost efficiencies from their technology investments.

One of the most notable shifts highlighted in the research involves corporate software spending. Nearly 33 percent of survey participants indicated that their organizations have scaled back traditional third-party software procurement in favor of building custom internal applications. This transition is largely being powered by agentic coding tools, which allow enterprise teams to develop specialized solutions in-house rather than relying on external software vendors.

Scale remains a determining factor in how successfully companies deploy advanced automation. Among enterprise organizations with annual revenues exceeding $1 billion, 40 percent reported actively scaling AI agents across their operations. That figure reflects a significant increase from the 27 percent recorded in McKinsey's previous annual study. Conversely, smaller enterprises have struggled to maintain pace, with only 22 percent reporting active scaling efforts—a metric that has remained largely unchanged year over year.

The discrepancy highlights how resource availability and infrastructure maturity favor larger institutions when expanding agentic workloads. However, corporate sentiment regarding the transformative potential of artificial intelligence remains bullish across the board. In its findings, McKinsey noted that organizational conviction in the technology is expanding at a faster rate than the tangible financial yields currently being logged on corporate balance sheets.

"Organizations’ conviction in AI is growing faster than the immediate financial returns they can attribute to it," McKinsey said in the report. "More expect AI to reshape their business over the next three years than did a year ago, and they continue to plan to invest more."

Despite high executive confidence, direct bottom-line gains remain modest for most enterprises. Only 6 percent of respondents reported that artificial intelligence contributes 5 percent or more to their company's earnings before interest and taxes (EBIT). While that figure has remained largely stagnant compared to previous survey periods, analysts view the stability as a baseline as organizations move beyond initial experimentation into broader deployment phases.

Operating expenses are emerging as a primary barrier to wider execution. Approximately 20 percent of surveyed organizations reported curbing their use of artificial intelligence due to high token costs and computational overhead. Even so, individual end-users are reporting tangible gains: 80 percent of respondents cited improvements in individual worker productivity, while 50 percent reported enhancements in executive decision-making processes.

The broader integration of these automated tools is also altering workforce forecasts. According to the report, 39 percent of respondents anticipate a reduction in total headcount over the coming 12 months as a direct result of AI implementation, up from 32 percent in the prior survey. Meanwhile, 43 percent of participants expect their organization's staffing levels to remain unaffected by technology deployments during the same period.