AQU Catalonia: When the Quality Agency Owns the Outcome Data

Bottom line up front: In Catalonia, the quality-assurance agency itself owns one of Europe's most comprehensive graduate-employment datasets — a triennial survey of nearly 49,000 graduates, conducted three years after graduation, measuring salary, skills match and contract quality. That changes what your internal course evaluation is for. When the regulator already holds rich longitudinal outcome data, a within-course satisfaction average is not a competing measure — it is a leading indicator that must connect to that outcome picture, or it tells the system nothing. AQU Catalunya is a preview of where evidence-led quality assurance is heading, and a satisfaction mean docks onto none of it.

A decentralised system most people outside Spain miss

Spain does not run quality assurance from a single national agency. The national body, ANECA, coexists with autonomous-community agencies, coordinated through REACU, the Spanish network of university-quality agencies — currently ANECA plus ten regional agencies across Andalusia, the Canary Islands, Aragón, Castilla y León, Galicia, Catalonia, the Balearic Islands, Valencia, Madrid and the Basque Country. For degrees delivered in Catalonia, AQU Catalunya has full legal authority to issue the ex-ante accreditation reports that validate programmes — a regional agency doing the substantive gatekeeping ANECA does elsewhere.

AQU is not a minor player. Founded as a consortium on 29 October 1996, it was the first quality-assurance agency in Spain; it has been a full member of ENQA since 2000, has been on the European Quality Assurance Register (EQAR) since 2008 with a status of "substantial compliance with the ESG," and was the first European QA agency to gain ISO 9001 certification. Its remit spans programme validation, institutional accreditation, teaching-staff evaluation and quality monitoring (AQU, Who we are).

The distinctive part: the agency owns the outcome data

What sets AQU apart from most national agencies is that it runs, jointly with the Catalan universities, a large triennial graduate-employment survey (the enquesta d'inserció laboral). This is not a light touch exercise. In the 2023 wave, 40,862 graduates responded, a 45.8% response rate, covering bachelor's, master's and doctoral graduates. The 2026 wave gathered around 49,000 respondents, surveying graduates three years after they finished and measuring salary, whether graduates perform university-level functions (a skills-match indicator), and contract type — which AQU describes as "one of the most comprehensive of its kind in Europe."

Sit with what that means. The regulator holds a longitudinal, skills-matched, whole-system view of what happened to graduates three years out. Against that, a course-level satisfaction mean collected in week twelve is not just less rigorous — it is a different kind of object entirely. It is cross-sectional, not longitudinal; it measures momentary reaction, not outcomes; and its raw average is not anchored to anything the outcome survey can use. The satisfaction number and the outcome dataset do not talk to each other.

The lesson: in-course evaluation must become a leading indicator

The productive response is not to compete with the agency's survey but to make your internal course evaluation the leading indicator that connects to it. AQU's own architecture points the way. Since 2007 its AUDIT programme has helped universities design internal quality-assurance systems (the SGIQ/IQAS), and since 2014 AQU certifies whether those systems actually run. Student voice enters accreditation through formal student reports, and programmes are re-accredited on a six-year cycle (four years for master's degrees). In other words, the system already expects your internal evaluation to be evidence of a working improvement loop — not a headline number.

That reframes the design brief. Your course evaluation should produce evidence that:

Points forward to outcomes, using the same constructs the outcome survey cares about (skills match, applied capability), so the two ends of the graduate lifecycle connect rather than sit in separate silos.
Demonstrates a closed loop, because that is literally what AUDIT/IQAS certification inspects — the action gap between collecting feedback and using it is what fails a review, not a mediocre mean.
Is comparable and legible, because a raw satisfaction average is not the interval-level, outcome-anchored evidence that an outcome-owning agency can integrate.

This is the same direction of travel as EUROGRADUATE at EU level and the graduate-tracer tradition across Europe: outcome data is becoming systematic and agency-owned, and in-course evaluation has to earn its place beside it.

But isn't this just a niche Spanish quirk?

The strongest objection is that a regional agency running a graduate survey is an idiosyncrasy of Spanish federalism, irrelevant to institutions elsewhere. Two reasons it is not. First, the model — the quality regulator owning longitudinal outcome data — is spreading, from EUROGRADUATE to national graduate-outcome surveys across the sector; AQU is early, not unique. Second, and more importantly, the argument does not depend on your regulator running a survey. It depends on a general truth: the moment outcome data exists anywhere in your quality ecosystem, a raw satisfaction mean looks threadbare by comparison, and the pressure to make in-course evaluation outcome-legible follows. AQU simply makes that future visible now.

A fairer objection is methodological: a survey three years post-graduation cannot cleanly attribute a graduate's salary to any single course, because too much intervenes. That is correct — and it is precisely why in-course evaluation (a leading, attributable signal about teaching and learning) and the outcome survey (a lagging, system-level signal about destinations) are complements, not substitutes. The mistake is to run the first as if the second did not exist. Neither can do the other's job; used together they triangulate what a single number never could.

Where Koji fits

Koji is an AI-native course-evaluation platform designed for exactly this leading-indicator role. Its AI-moderated conversational interviews probe beyond a rating to capture applied capability and skills-match signals in students' own words — the constructs an outcome survey later measures at destination. Automatic thematic analysis turns open text into structured, comparable themes; quality scoring keeps the evidence trustworthy; and programme- and institution-level reporting produces the closed-loop trail that an AUDIT/IQAS-style certification inspects. Formative mid-cycle collection and closing-the-loop action tracking demonstrate that feedback changed something — the evidence a six-year accreditation review actually wants — and data handling is GDPR/AVG-compliant and EU-appropriate. Because the AI moderation is standardized, comparisons across cohorts and sites are not distorted by inconsistent human interviewers.

Koji does not replace an agency's graduate survey or claim to measure destinations — it surfaces the in-course, outcome-oriented evidence that connects to that survey, and mitigates the illegibility of a raw satisfaction mean. Institutions running wider stakeholder or alumni research often use the same AI interview engine on the main Koji platform.

If your regulator is moving towards owning outcome data — and in Catalonia it already does — the question is whether your course evaluation can dock onto it. Explore Koji for Education to build evaluation that connects the in-course signal to the outcome picture.

Frequently asked questions

What is AQU Catalunya? AQU Catalunya (Agència per a la Qualitat del Sistema Universitari de Catalunya) is the quality-assurance agency for Catalonia's university system. Founded in 1996 as the first QA agency in Spain, it is a full ENQA member, EQAR-registered, and handles programme validation, accreditation, teaching-staff evaluation and quality monitoring for Catalan universities.

How can a regional agency accredit degrees if Spain has a national agency? Spain's quality assurance is decentralised. ANECA operates nationally, but autonomous communities have their own EQAR-registered agencies coordinated through the REACU network. For degrees delivered in Catalonia, AQU has the legal authority to issue the ex-ante accreditation reports that validate programmes.

What makes AQU's graduate-employment survey notable? Its scale and ownership. AQU and the Catalan universities run a triennial survey — around 49,000 respondents in the 2026 wave — three years after graduation, measuring salary, skills match and contract quality. AQU calls it one of the most comprehensive of its kind in Europe, and it means the regulator itself holds longitudinal outcome data.

Does that make internal course evaluation redundant? No. A survey three years out cannot attribute outcomes to a single course, and it collects nothing during the course. In-course evaluation is the leading, attributable signal; the outcome survey is the lagging, system-level one. They are complements — the mistake is running course evaluation as if the outcome data did not exist.

What should institutions take from the AQU model? That in-course evaluation must become outcome-legible: aligned to skills-match constructs, demonstrating a closed improvement loop, and comparable enough to sit beside agency-owned outcome data. A raw satisfaction average meets none of those tests.