There is a specific kind of institutional waste that rarely appears in budget reviews: the expenditure of skilled human attention on tasks that are, by nature, mechanical. In university assessment operations, this waste is structural. It is built into the way most institutions have organised their exam teams, and it persists not because administrators are unaware of automation, but because the gap between what software could handle and what it actually handles in a given institution has never been made explicit enough to force a decision.
The consequences surface quietly. An assessment coordinator spends three weeks before each examination period manually reconciling candidate enrolments, room allocations, and accessibility requirements across systems that do not speak to each other. A senior officer reviews each exam paper against a formal quality checklist (checking font sizes, Bloom's taxonomy coverage, and marking scheme alignment), tasks that an automated system could resolve in seconds. A results team exports data from one platform, reformats it in a spreadsheet, and re imports it into a reporting system, introducing transcription risk at every handoff. None of these individuals was hired to do this work. All of them are doing it anyway.
The Misallocation is Documented, Not Anecdotal
In a survey conducted by EDUCAUSE, 2024 Core Data Service, spanning 312 higher education institutions globally, 68% of respondents reported administrative time lost to manual data re entry, and 72% flagged accreditation data preparation as dominated by manual processes. To put that plainly: more than two in three universities have qualified staff copying data between systems by hand, and nearly three in four are preparing high stakes compliance records the same way. These are not marginal inefficiencies; they are a structural diversion of expert attention toward work that generates no analytical value and introduces real error risk every time a human touches data that software could handle cleanly.
Modern Campus, in its 2025 institutional analysis What Are Academic Operations in Higher Ed?, observed that many institutions are still running critical academic operations through manual, outdated processes that drain resources and frustrate both staff and students. In exam operations specifically, that coordination surface is especially wide: question authoring, moderation, scheduling, invigilation management, candidate communications, results processing, and certification all involve handoffs that, in most institutions, require human intervention at every junction.
Researchers publishing in Springer Nature titled Transforming Higher Education with Robotic Process Automation reviewed 54 studies on automation in education between 2020 and 2024, finding that robotic process automation has the potential to handle up to 70% of repetitive tasks in the sector and that institutions which implemented it saw measurable gains in time savings, error reduction, and staff satisfaction. The gap between that potential and current practice is precisely where institutional cost accumulates, and where staff capacity quietly drains.
The Exam Paper Review Problem
One of the more instructive examples of misallocated human effort sits at the beginning of the assessment lifecycle: exam paper quality assurance. In most universities, this is a manual process. Senior academics or assessment officers review papers against formal specifications (question format compliance, Bloom's cognitive level distribution, coverage weighting) before papers are approved for printing or digital delivery.
Researchers writing in Springer's Education and Information Technologies journal published a study titled Classification Driven Intelligent System for Automated Evaluation of Higher Education Exam Paper Quality in 2024, noting that the prevailing manual review process is tedious, lengthy, and in some cases inconsistent, often because human reviewers focus only on formal specifications while content level analysis goes underweighted. Their automated system, which categorised questions by Bloom's cognitive levels and checked formal compliance simultaneously, achieved classification accuracy that exceeded manual review on consistency grounds.
Doaa Mohamed Elbourhamy of Kafrelshiekh University published a study in PeerJ Computer Science titled Automated Evaluation Systems to Enhance Exam Quality and Reduce Test Anxiety in February 2025, analysing 1,800 questions across 30 university exam papers. The research found that adherence to basic structural and technical standards was missing in a significant portion of papers reviewed, and that an automated evaluation system raised exam paper quality substantially while reducing academic stress among students. The implication is not simply operational; poorly constructed exam papers carry downstream consequences for student outcomes and institutional accreditation standing.
Where Purpose Built Infrastructure Changes the Calculus
The distinction worth making here is between general automation, applying scripted tools to tasks that were designed for human execution and purpose built assessment infrastructure, in which workflow logic is embedded in the platform architecture itself. The former reduces friction at specific handoff points. The latter eliminates the handoffs.
Platforms designed around the full assessment lifecycle, Janison's assessment platform being one example, embed coordination logic into the workflow rather than leaving it to administrators to manage manually. Question banking with role based authoring permissions, automated moderation routing, integrated scheduling, candidate management, and results reporting exist within a single environment rather than across a constellation of exports and imports. The coordinator's role shifts from managing data movement to managing exceptions.
This matters because the cost of fragmentation is not just time it is attention. Assessment professionals hired for their expertise in psychometrics, equity, and quality assurance are spending cycles on logistics that should be invisible to them. EDUCAUSE, in its 2025 Technology Budgets and Staffing QuickPoll Results, found respondents expressing overwhelming optimism about automation's potential to offload administrative burdens and mundane tasks, with 83% of institutions considering, planning, or implementing technology driven operational streamlining. Optimism, though, is not the same as execution, and the gap between intention and implementation in assessment operations remains wide at most institutions.
The Staff Wellbeing Dimension Institutions Undercount
There is a personnel cost that rarely appears in technology ROI calculations. A research team from the California University of Science and Medicine published a study in Frontiers in Psychiatry titled The Overlooked Pillars of Medical Education: Addressing the Mental Health and Well Being of Medical School Administrative Staff in November 2025, finding that administrative staff are vulnerable to systemic disruptions due to their vital but often invisible role in maintaining institutional continuity, and that their mental health burden during periods of operational pressure rivals that of students and faculty yet this workforce remains largely absent from institutional support discussions.
Research.com's compilation of workforce analytics, published as Teacher Burnout Statistics for 2026: Challenges in K 12 and Higher Education, drawing on Gallup's 2025 Workplace Insights data, found that burnout among university educators reached 38%, placing university teaching among the top two occupations by burnout rate in the US workforce. Exam operations staff, sitting adjacent to faculty but rarely counted in faculty wellbeing surveys, absorb a disproportionate share of peak period operational load, precisely the load that automation is best positioned to absorb instead.
EDUCAUSE placed administrative simplification second on its 2025 Top 10 Institutional Technology Agenda, a ranking that reflects how broadly the sector has come to recognise the problem. Recognition, however, has not yet translated into action at the operational level where assessment teams actually work.
The more honest institutional question is not whether automation is available, because it is. It is whether leadership has been willing to map the full surface of what their exam team actually does each cycle, and confront how much of it software should have absorbed years ago. Every hour a qualified assessment professional spends chasing a spreadsheet discrepancy or manually routing a moderation email is an hour not spent on the work that credentials are built on. That substitution has a cost. Most institutions have simply never been made to count it.
