Leaders Say AI Saves Time. Workers Are Not So Sure

Your daily brief on leadership and business

Ask the boss whether AI is making the company more productive and you are likely to hear “yes”. Ask the people doing the work and the answer gets noticeably quieter. That is the headline finding of new research from workforce technology company Vertage, which surveyed 800 enterprise decision-makers and 1,200 workers across six markets. The study was commissioned by the company itself.

A 28-point view from the top

According to the survey, 71% of decision-makers said AI had increased their organisation’s capacity. Among workers, only 43% agreed. That is a 28-percentage-point gap, wide enough to park a whole AI roadmap in. The gap does not by itself prove that AI has failed to raise productivity: leaders and employees may simply be measuring capacity in different ways. But it does suggest that the time saved on a slide in the boardroom does not always show up on an employee’s calendar.

The study, titled The Outcome Economy at Work, argues that buying AI is no longer the hard part. Companies have to decide which activities should change, who is accountable for results and how people and automated systems will work together. While 74% of enterprises say they define significant work around outcomes rather than roles or headcount, only 15% meet the study’s criteria for advanced “outcome readiness”. Firms in that advanced group reported 91% customer satisfaction versus 67% for others, and 86% delivery on strategic goals versus 65%, though these are correlations rather than proof of cause.

“AI adoption no longer separates strong businesses from the rest. Outcome readiness does,” said Edzard Overbeek, Vertage’s group CEO.

Where the saved time goes

The practical lesson is to ask what has actually changed in the work. Saved minutes can easily be swallowed by checking AI output, extra reporting or workloads that simply never shrank. Managers need to be clear about who reviews AI-generated results, who handles exceptions and who remains accountable. Measuring service quality, workload, errors and turnaround time alongside adoption figures gives a more honest picture than counting licences, and it might even help both sides of the 28-point gap agree on what “productive” means.