AI Productivity: Count the Rework and the Impact on Engagement
Research Highlights Costs at the Handoff
Six Recommendations for AI Pilots
Connect AI Measurement to Engagement
How much time are we truly saving with AI when its output almost always has to be reviewed? How can organizations better understand how much time they are saving and what can they do to enhance productivity? These are questions this paper attempts to help answer.An employee uses artificial intelligence to finish a customer summary in four minutes instead of 20. A colleague then spends half an hour checking its conclusions. The first employee’s dashboard shows a productivity gain, while the organization absorbs additional work when the extra review time is added.
This hypothetical example illustrates a measurement problem identified by Gleb Tsipursky, PhD, behavioral scientist, CEO of Disaster Avoidance Experts, and former Assistant Professor of history at the Ohio State University’s Newark campus. In his submitted commentary, “An AI Time-Saving Claim Needs a Second Column for Rework,” he argues that organizations should follow AI-assisted work through the complete process before claiming a productivity improvement.For enterprise engagement practitioners, the implications extend to trust, collaboration, learning, and recognition. When one department receives credit for faster work while another absorbs the corrections, performance measures can conceal the burdens employees experience.
Research Highlights Costs at the Handoff
A BetterUp Labs and Stanford Social Media Lab study, based on a September 2025 online survey of 1,150 full-time US desk workers, found that 40% reported receiving AI-generated “workslop” in the previous month. The term describes material that appears polished but lacks the substance needed to advance a task. Recipients reported spending about two hours resolving each incident. These are reported experiences, rather than independently measured productivity losses across all employers. Nevertheless, they highlight why measuring
Employee judgment presents another consideration. A 2025 CHI study led by Hao-Ping (Hank) Lee and coauthors, summarized by Microsoft Research, surveyed 319 knowledge workers who provided 936 examples of generative AI use. Higher confidence in AI was associated with less reported critical thinking. The study does not establish that AI causes a decline in capability but raises questions about how organizations preserve opportunities to evaluate evidence and challenge recommendations.
A complementary Qualtrics report, Experience Is the Future of Business, argues that faster decisions across individual functions require coordination to produce consistent employee and customer experiences.
Six Recommendations for AI Pilots
Tsipursky proposes several steps for determining whether AI improves the complete outcome:- Track downstream work. Record the creator’s time alongside the verification, clarification, correction, and repeat work required by recipients. Attribute additional effort to AI only when the connection can be documented.
- Start with a representative sample. Follow a few dozen work items through their handoffs, asking receiving teams to record the effort required to make the output useful. Compare similar work completed without AI where possible.
- Identify who absorbs the burden. Examine whether apparent gains in one department depend on additional work in another. Make those costs visible when reviewing budgets, performance, and recognition.
- Measure the experience. Ask whether recipients understand the answer, know who owns the next decision, and can resolve a need without repeating information. Follow a service request from intake to resolution to identify weak handoffs.
- Check independent judgment. Ask employees to explain an important AI-assisted recommendation in their own words, identify an exception the system might miss, and describe evidence that would change their conclusion.
- Reward complete outcomes. Recognize improvements that help colleagues and customers achieve their goals, including clearer information, fewer corrections, and better service without oversight or problem correction time exceeding the initial time savings.
Connect AI Measurement to Engagement
The engagement opportunity is to involve both the people producing AI-assisted work and those receiving it in defining success. Their feedback can reveal missing information, unclear responsibilities, and service problems that a task-speed measure overlooks. Tsipursky recommends keeping measures visible to both sides of each handoff and revising them as employees identify previously uncounted work. This gives teams a practical role in improving adoption and helps leaders align recognition with the organization’s broader goals.An AI initiative may save only a few minutes yet deliver substantial value if it reduces confusion, improves service, and helps employees exercise sound judgment. Measuring those outcomes provides a stronger basis for deciding which applications deserve wider use.
Enterprise Engagement Alliance Services
Founded in 2009, the Enterprise Engagement Alliance helps organizations create measurable value through people by aligning employees, customers, partners, and other stakeholders around a shared purpose, goals, and values. Explore EEA services and membership options.
1. News, research, and marketing opportunities
-
ESM Weekly: Stakeholder management and enterprise engagement coverage since 2009. Subscribe | Media kit.
-
RRN Weekly: Total rewards and recognition coverage since 1996. Subscribe | Media kit.
-
EEA YouTube channel: Enterprise engagement, human capital, and total rewards programs since 2020.

2. Membership, learning, and certification
The EEA Impact Academy teaches a systematic approach to aligning people, purpose, and performance. Complimentary Professional Management Membership is available to qualified corporate executives and managers, professors, and educators. Paid Corporate and Solution Provider memberships support team learning and professional development. Explore learning and registration.
3. Books and implementation resources
Enterprise Engagement for CEOs introduces the framework for senior leaders.
Enterprise Engagement: The Roadmap provides a practical implementation framework.
4. Advisory services, assessments, and impact analysis
Advisory services and research support stakeholder management and implementation. EEA resources include the Enterprise Engagement Index, People Value Impact Calculator, and independent analysis of engagement, incentive, recognition, loyalty, and related initiatives.
5. Permission-based targeted business development
Strategic communications and business development services help identify and build relationships with the people most likely to buy through combined newsletters, thought leadership, and social media.
Contact: Bruce Bolger at TheICEE.org; 914-591-7600, ext. 230.














