When a slow process becomes a large five-figure cost

Late exam grading was a known irritation, but no one had connected it to the downstream load it created on student services, or put a number against it. The question was whether a slow back-office process was quietly generating a real operational cost.

The client is a large online higher-education provider serving a high volume of students and running a correspondingly large number of exams each year. Grading delays were treated as an academic-operations issue, handled within the teaching workflow.

The student-services team, meanwhile, absorbed a steady stream of complaints without anyone establishing whether the two were connected. The link between a slow process and a rising support cost had never been tested — so the true cost of late grading stayed invisible.

Newly available exam-attempt and student-incident data let us join the two for the first time. We compared the complaint rate on exams graded on time against those graded late, measured how that rate escalated with each additional week of delay, and looked for behavioural precursors — early signals that a complaint was coming.

We then extrapolated the relationship across a full year of exam volume to size the annual operational cost, using a deliberately conservative per-incident effort estimate.

Exams graded on time generated almost no complaints. Late-graded exams generated complaints at 7.4%, climbing to around 13% once grading was three or more weeks late. The most useful finding was a behavioural early warning: around half of all complaints were preceded by the student checking their grade in the portal within 24 hours — a visible signal that intervention could get ahead of.

For the first time, the business had a defensible price tag on a process problem it had been treating as minor. That reframed late grading from an academic-workflow inconvenience into a quantified operational cost with a clear owner.

It also surfaced a practical lever: the in-portal grade-check gave student services an early-warning signal to intervene before a complaint was raised. The analysis didn't fix the grading process; it proved the process was worth fixing, and showed where to act first.

Operational Diagnosis is one of five Hira Deep-Dive Sprints — a focused, four-week engagement that traces an operational problem through to its commercial cost and finds the points where intervention pays off. It's the right fit when a process feels slow or painful but nobody has costed what it's actually doing downstream.