The Graduate-Outcome Lag: Why Employability Data Arrives Too Late to Fix a Course

Bottom line up front: The dominant measure of graduate employability in European higher education is a lagging indicator — it tells you what happened to students long after the course that shaped them is over. The UK's Graduate Outcomes survey, for instance, contacts graduates roughly 15 months after they leave. That is invaluable for accountability and useless for improvement: by the time the data lands, the curriculum, the cohort and often the teaching team have moved on. If programmes want to act on employability rather than merely report it, they need leading indicators — and a well-designed course evaluation is one of the few they already control.

Lagging indicators: essential, but always looking backwards

Graduate-outcome surveys are among the most important datasets in the sector. The UK's Graduate Outcomes survey is the largest annual social survey in the country; the most recent HESA release reports that 88% of 2022/23 graduates were in work or further study at the 15-month census point. Comparable national instruments exist across Europe, and graduate tracer studies extend the picture further out.

But every one of these shares a structural limitation: they are lagging indicators, borrowed from the language of performance measurement. A lagging indicator confirms an outcome after it has occurred. It is accountable, auditable and, for the purpose of steering a live programme, arrives too late to change anything. Three compounding delays make this acute:

1. The census lag. Outcomes are measured 15 months post-graduation.
2. The programme lag. For a three-year degree, the teaching being judged happened up to four-plus years before the data appears.
3. The response lag. Graduate surveys rely on tracing alumni, with response rates around 39% for the UK Graduate Outcomes survey — introducing exactly the non-response bias that makes acting on the tail of the distribution hazardous.

By the time a programme director learns that a cohort struggled to convert their degree into graduate employment, the students who could benefit from the fix have already graduated. The data is a post-mortem, not a diagnosis.

The case for leading indicators

A leading indicator measures something now that predicts the outcome later — and can be acted on while the outcome is still changeable. In employability terms, the lagging indicator is "did graduates get relevant jobs?" The leading indicators are the things that produce that result: are students developing the skills employers actually need? Do they perceive the curriculum as building capability, not just knowledge? Are they experiencing the applied, integrative learning that predicts graduate success?

Course evaluation is the natural home for these questions — if it is designed to ask them. Most SET instruments are not. They ask about satisfaction, clarity and workload, and stop. They almost never ask whether a course is building the competences a fast-moving labour market demands. And that market is moving fast: the World Economic Forum's Future of Jobs Report 2025 estimates that 39% of workers' core skills will change by 2030. A course evaluation that measures only satisfaction is blind to the single most important thing a programme needs to steer: whether it is keeping pace with that skills shift while students are still enrolled.

This reframes the purpose of evaluation. Instead of a rear-view satisfaction survey, a course evaluation can be an early-warning system for employability — surfacing, mid-programme, whether students are gaining the applied and transferable skills that graduate-outcome surveys will only confirm or refute years later. We have argued before that evaluation should move beyond satisfaction to measuring skills; the leading-indicator frame is why that shift matters operationally.

But can students really assess their own employability?

The strongest objection is a validity one: students are poor judges of their own future employability, so asking them is just satisfaction data in a new costume. There is real substance here. Perceived skill is not demonstrated skill, and self-report is vulnerable to the Dunning-Kruger dynamics that dog all self-assessment. A course evaluation cannot measure employability the way a graduate survey can.

The response is to be precise about what a leading indicator claims. It does not claim to measure the outcome; it claims to measure a predictor you can act on. Students may not know whether they are employable, but they are excellent witnesses to whether a course gave them chances to apply knowledge, work in teams, tackle authentic problems, or build the specific capabilities a discipline requires — the proximal experiences that mediate later outcomes. The trick is to stop asking students to grade their employability and start asking them to describe their learning experience in ways that map to known predictors of it. And crucially, the leading indicator does not stand alone: it is validated against the lagging graduate-outcome data over time, so the two form a loop rather than a substitution. The course evaluation gives you something to act on now; the graduate survey tells you, later, whether you were acting on the right things.

Building the loop: from evaluation to employability

A leading-indicator approach to employability changes what a programme measures and when:

• Ask about applied and transferable learning, not just satisfaction. Whether students had authentic, integrative, skill-building experiences is a predictor you can influence next term.
• Collect it mid-cycle, not only at the end. A leading indicator is worthless if it arrives after the cohort leaves. Formative, mid-programme collection is what makes it actionable.
• Map evaluation to a skills framework. Aligning questions to a competence taxonomy — as in ESCO-mapped evaluation — connects what students report to what employers name.
• Close the loop with the lagging data. Validate your leading indicators against graduate-outcome results over time, keeping the ones that actually predict, and feed employer signals back in through an employer feedback loop.

How Koji addresses the lag

Koji for Education is designed to turn course evaluation into a leading indicator rather than a rear-view mirror. Its AI-moderated conversational interviews can probe how students experienced a course — did they apply knowledge to authentic problems, work in teams, build specific capabilities — instead of extracting a single satisfaction number, which is exactly the proximal, actionable signal a graduate survey cannot capture until years later. Because collection is not tied to end-of-term, Koji supports formative, mid-cycle evaluation, so a programme can act on the signal while the cohort is still enrolled. Its six structured question types and automatic thematic analysis let institutions map open-ended student experience to a competence framework and track it programme-wide, and its closing-the-loop action tracking connects what evaluation surfaces to what the programme does next. Koji does not replace graduate-outcome surveys — the lagging indicator remains essential for accountability and validation — but it gives programmes the leading indicator they have been missing. The same interview engine powers skills and experience research on the main Koji platform.

Why accreditors increasingly want both

This is not only an internal-improvement argument. European quality assurance is moving toward evidence that programmes actively manage employability rather than merely report on it after the fact. Under the Standards and Guidelines for Quality Assurance in the European Higher Education Area (ESG), institutions are expected to monitor and periodically review the ongoing relevance of their programmes — a standard a lagging indicator alone cannot satisfy, because it offers nothing to act on within the review cycle. A leading indicator drawn from course evaluation gives accreditation panels something a graduate survey cannot: contemporaneous evidence that a programme is steering toward employability while students are still enrolled, closing the loop between what is taught, what students experience, and what the labour market rewards.

The takeaway

Graduate-outcome surveys will always be a lagging indicator; that is their job, and they do it well. But a sector that measures employability only after graduation has no steering wheel — only a rear-view mirror. Course evaluation, redesigned to ask about applied and transferable learning and collected while it can still change something, is the leading indicator hiding in plain sight. The question is not whether to keep the graduate surveys. It is whether to keep flying blind between them.

Want to turn course evaluation into an early-warning system for employability? Explore Koji for Education.