A review of 129 academic studies finds that hotel robot deployments routinely hit their operational targets while quietly hollowing out the staff engagement that makes those numbers sustainable, raising turnover risk for operators who read dashboard gains as proof the rollout worked.

The study, "Hybrid Service Agency: Human–Robot Collaboration and Workforce Transformation in Hospitality and Tourism" by M. Borghi, P. Ratcharak and B. Taheri, appears in the 2026 volume of The Service Industries Journal.

It synthesizes more than a decade of research on service robots in hospitality and tourism. The conclusion for operators is uncomfortable: room-service times and guest scores can move in the right direction while team members stop believing the work is worth doing.

What do the dashboards miss?

Borghi, Ratcharak and Taheri cite one recent hotel-employee study in which visible service performance rose even as staff passion for the job declined, once workers read the robot purchase as a signal that management valued efficiency over people. The result is the most expensive kind of success: one that shows up in the weekly KPI report months before it surfaces in the resignation queue.

"The robot is the easy part of the investment," the authors write. "The team around it decides whether it pays off."

A second strand of evidence links job-insecurity perceptions to active sabotage. Staff who saw robots as a threat to their roles were more likely to mistreat them. A robot that keeps "breaking down" on a single shift may tell the operator more about the team on that shift than about the hardware.

Why do rollouts lose staff before they lose guests?

The review identifies two recurring breakdowns:

• Unclear intent. Managers frame robots as support for stretched teams. Without explanation, the same purchase reads to staff as a step toward reduced hours.
• Unclear roles. When the split between robot-owned and human-owned tasks is ambiguous, willingness to work with the robot drops regardless of any general openness to automation.

Robot design shapes the threat read. Back-of-house units handling repetitive or physical work drew the least resistance. Socially capable models triggered the most insecurity, because they look close enough to perform the parts of the job workers value most.

A young workforce is no buffer. Generation Z employees accept robots in principle but show heightened sensitivity to the fairness of the rollout and to how much autonomy they retain. Once insecurity takes hold, it translates directly into stated intentions to quit.

What can operators change?

Across the studies, supportive and transformational leadership consistently weakened the link between feeling threatened by robots and pulling back from work. The evidence base is still maturing, but the pattern is consistent enough to act on, the authors argue.

A four-step playbook runs through the literature:

• State the why before launch. Tell the team what the robot is for, what it means for hours, and which roles change. Rumour costs more than honest news.
• Define task ownership in writing. Spell out which tasks belong to the robot, which stay with people, and when guests should ask for a human. Give staff authority to override the machine.
• Train after purpose and roles are clear. Let supervisors help staff reshape their own tasks around the robot, and recognize managers who build genuine rapport with their teams.
• Track more than speed. Add engagement scores, voluntary turnover, and peer-to-peer problem-sharing to the dashboard. Treat dips as a signal to revisit the management plan, not as staff resistance.

Where does the upside actually sit?

The strongest-performing teams in the reviewed studies split work by comparative advantage. Robots take the consistent, repetitive, and physically demanding tasks at clearly defined touchpoints; guest satisfaction lands on par with fully human service. People own judgement, reassurance, quality checks, and anything unexpected.

When a robot stalls in a corridor or delivers the wrong order, employees with the autonomy to step in adapt; those waiting on instructions tend to push back. The researchers term this team-level capacity "hybrid service agency," the joint output people and robots can produce that neither can alone.

For operators planning 2026 capex on robotics, the cost line that decides payback is labor. The studies reviewed suggest a robot program without a parallel investment in communication, role clarity, and supervisory practice will buy speed today and pay for it in replacement-hiring costs later.