From the Campus Operations edition of October 20, 2026
Hague University of Applied Sciences study of 159 staff finds GenAI made tasks faster in a randomized trial while perceived usefulness fell over four weeks, most sharply among support staff
The Hague University of Applied Sciences (Garg, Kernan Freire, Bruck, Tanjung, Bakker) · Public university of applied sciences: staff use of an institutional GenAI platform · The Hague, Netherlands
- Publisher
- SSRN
- Original publication
- Posted 6 October 2026 (SSRN preprint, 31 pages)
- Source retrieved
- 2026-10-21
What happened
In a four-week pilot, 159 staff across education, research and support received an institutional GenAI platform. In a micro-randomized trial, staff with GenAI completed a work assignment faster and automated evaluators rated their outputs higher, but their own quality ratings were no higher than those without GenAI. Perceived usefulness fell across all three domains, from 5.30 to 4.20 on a 7-point scale. Researchers showed the largest productivity gains; support staff showed the sharpest decline in perceived usefulness. The authors call for longitudinal evaluation and domain-specific support.
Why it matters
One of few causal studies of GenAI in university staff work, and it cuts against the satisfaction-survey evidence most campus rollouts rely on. Support staff, the core of campus operations, lost enthusiasm fastest. Older than the last run but not archived; included now because this week's rollouts at Stanford, UBC and CSU report access, not outcomes.
Evidence and measured results
Mixed-methods design with surveys, interviews and a micro-RCT on one assignment; n=159; four weeks. Only the abstract was inspected; effect sizes for speed and quality are not given there. Automated quality ratings may weigh different dimensions than human ratings, as the authors note.
Limitations and uncertainty
Preprint; abstract only inspected; single institution in the Netherlands; four weeks; one assignment in the RCT; automated evaluators may not match human judgment. Transferability to U.S. campuses uncertain.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-20; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Campuses justify staff AI licenses with early satisfaction, but this trial shows perceived usefulness falling while measured speed improved.
- Stakeholders
- CIO and AI program lead
- HR and staff development
- Administrative unit managers
- Institutional research or evaluation office
- Discovery questions
- How is value measured beyond launch surveys?
- Which support-staff tasks are targeted?
- Will perceptions be tracked over several months?
- Who reviews output quality: people or automated scoring?
- Value hypothesis
- Interpretation: a longitudinal evaluation could prevent over-buying for roles where value fades; the study is small and short.
- Potential engagement
- A bounded three-month evaluation design for one administrative staff group combining task timing, human quality review and perception surveys.
- Unsupported claims
- Specific productivity percentages, not given in the abstract
- That results generalize to U.S. institutions
- That GenAI improves support-staff quality as judged by humans
- Guaranteed savings
Pre-sales engineering
Role takeaway
- Fit
- Evidence about task-technology fit rather than platform architecture.
- Architecture and integration
- Institutional GenAI platform, unspecified; micro-RCT on one work assignment.
- Prerequisites
- Defined task set per staff domain
- Human quality rubric
- Usage telemetry over time
- Constraints
- Four-week window
- Single Dutch institution
- Abstract-only inspection
- Security
- Not addressed in the abstract.
- Proof of value
- Compare automated and human quality ratings on local tasks
- Measure task time at baseline and after months of use
- Track perceived usefulness by role over time
Delivery
Role takeaway
- Work
- Build a longitudinal evaluation into staff AI rollouts, with domain-specific support for operations staff.
- Dependencies
- Willing pilot units
- Survey instrument and timing
- Human reviewers for output quality
- Ownership
- The AI program office owns evaluation; unit managers own task selection.
- Skills and adoption
- Digital literacy and domain-specific support are named conditions; plan coaching for support staff.
- Governance checkpoints
- Baseline before access
- Review at four weeks and at three months
- Use-case adjustment before expansion
- Acceptance criteria
- Proposed: task time and human-rated quality measured at baseline and three months
- Proposed: perceived usefulness tracked by role at three points
- Proposed: expansion decision documented against these measures
- Risks
- Novelty effects mistaken for value
- Automated metrics overstate quality
- Support staff disengage without tailored use cases
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Limited direct relevance; the platform is described only as institutional. Task-technology fit by domain matters more than tool choice in this evidence.
Governance
Who approves, reviews and stays accountable for outcomes?
Supports requiring longitudinal evaluation rather than early surveys before scaling licenses; domain-specific evaluation for support staff in particular.
Security and privacy
What data, permissions and controls need testing?
Not addressed in the abstract.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Support staff showed the sharpest decline in perceived usefulness; digital literacy and domain-specific support named as adoption conditions.
Procurement
What should contracts, pricing and exit terms secure?
Renewal decisions based on early satisfaction may overstate durable value; the study suggests measuring perceptions and performance over time.
Operating model
Which teams own the service once it runs?
AI adoption programs for operations staff need task-level fit analysis and ongoing support, not one-off launch training.
Publication history
- 2026-10-20Campus Operations · Issue 435 resources
Stable resource ID: hague-uas-genai-staff-rct-perception-gap-20261006