Course it’s not the careers service

Three graduates, speaking in focus groups commissioned by the Office for Students (OfS), sum it up:

There was nothing industry-specific, there was nothing course-specific. There was nothing person-specific. It was all just like generic CV tips.”

That’s a social sciences graduate. A law student put it differently:

It was kind of a steep learning curve to realise that the support services were there, but if you wanted to access them, you had to ask. No one was really going to check in on you.”

And an arts and humanities graduate reflecting on the job market:

You understand it is competitive, but you don’t know the nitty gritty of why it’s competitive, which kind of gives you hope until you actually have to face it.”

These graduates are precise about what went wrong – nobody told them where their degree leads, the support that existed was generic and invisible, and the gap between what they expected and what they encountered after graduating was a shock that nobody had prepared them for.

OfS commissioned the research, read the transcripts and published the findings. But its headline response is that institutions should consider making their careers services more visible. Students are diagnosing a student experience design problem. The regulator is prescribing a communications fix.

OfS has published two documents on what it’s calling “preparation for the transition out of higher education.” The first is a 40-page research report by IFF Research, based on a survey of 1,671 recent graduates and three focus groups involving 18 participants.

The second is a 12-page “student insight report” authored by OfS itself, which digests the IFF findings and wraps them in contextual framing from the Graduate Futures Institute (GFI) – formerly the Association of Graduate Careers Advisory Services (AGCAS) – two institutional case studies, and a set of “points for universities and colleges to consider.”

The research covers graduates from the 2022/23 to 2024/25 academic years, surveyed retrospectively in September 2025.

Josh Fleming, OfS’s director of strategy and delivery, frames the research as showing that “many institutions are doing impressive work to support students to achieve their goals” while identifying “ways students’ experiences could be improved even further.” GFI’s Martin Edmondson uses the occasion to call for “appropriate institutional resourcing and prioritisation” of careers services.

The IFF research report itself is solid, methodical, and contains a wealth of granular data that will be useful to universities. OfS’s insight report that sits on top of it is something else – a carefully editorialised document that softens the more uncomfortable findings and steers the reader toward conclusions that are more palatable than the evidence warrants.

Unpicking the gap between the two is worth doing, not least because we’ll shortly be publishing our own analysis of career preparation data from our polling, drawn from over 20,000 responses – which tells a different story about where problems sit and what would actually fix them.

If there is one, the meta-narrative, across all of the evidence, is that the system keeps mishearing what students are telling it.

Half measures

The IFF research produces a picture that is, depending on your general outlook on life, either reassuringly adequate or viscerally alarming.

Half (50 per cent) of graduates surveyed felt prepared for life after graduation. A third (33 per cent) felt unprepared. 62 per cent felt confident about achieving their post-graduation goals, 88 per cent reported having received some form of support from their institution, 79 per cent were in employment at the time of the survey with 10 per cent unemployed, and only a third (33 per cent) had used their institution’s careers service.

The headline numbers are what they are, but the more interesting findings are in the subgroup analysis. Postgraduate students were more likely to feel prepared (61 per cent) than undergraduates (46 per cent). HE in FE graduates reported significantly higher preparedness (69 per cent) and confidence (80 per cent) than university graduates (50 per cent and 62 per cent respectively).

Unsurprisingly, graduates of vocational and science, technology, engineering, and mathematics (STEM) subjects consistently scored higher on preparedness, confidence, and satisfaction than those from arts, humanities, and social sciences.

Arts and humanities graduates were the least confident (47 per cent), the most likely to feel unprepared (43 per cent), and the most likely to identify lack of a professional network as a barrier (50 per cent).

Subject of study is, by a distance, the most powerful predictor in the dataset – a finding which our own forthcoming analysis of current student career confidence confirms with more statistical precision (what happens when you control for it is even more interesting).

But the IFF report goes further on several aspects that OfS’ insight report chooses to underplay, and it’s worth noting three of them.

Soft focus

There is a persistent and substantial disability gap running through the IFF data. Graduates with a long-term illness, health condition, or disability were around 10 percentage points less likely to agree that their HE experience had been helpful on nearly every dimension measured.

They were less satisfied with support (51 per cent versus 61 per cent), less likely to rate advice quality positively (58 per cent versus 65 per cent), and more likely to feel unprepared for finding work at their skill level (36 per cent versus 24 per cent).

The insight report mentions disability three times – once to note that disabled students could benefit from tailored support, once to reference an existing OfS insight brief, and once in the checklist at the end. The scale and consistency of the gap isn’t conveyed.

Our data, controlling for subject, finds a disability residual that persists even within the same courses – disabled students are overrepresented in the lowest career confidence group and underrepresented in the highest, and what they study doesn’t explain the difference.

The parental HE qualification gap is more extensive than the insight report conveys, too. The IFF data shows that graduates whose parents didn’t have a degree were less likely to access institutional support across a range of measures – careers services (30 per cent versus 37 per cent), departmental events (16 per cent versus 27 per cent), work experience (19 per cent versus 25 per cent), enrichment activities (16 per cent versus 25 per cent), and industry connections (14 per cent versus 19 per cent). When they did access formal support, they were less likely to find it useful (47 per cent versus 55 per cent).

The insight report frames the parental HE gap almost entirely around family support – noting that graduates with HE-qualified parents found family advice more useful – and proposes that institutions should “compensate” for absent family networks.

That’s the right instinct, but it sidesteps the more uncomfortable finding that first-generation students are also less likely to engage with and benefit from the institutional support that already exists. Some students lack family networks, yes – but the institutional alternative isn’t reaching the students who most need it either.

Then there’s the postgraduate satisfaction finding – 54 per cent, the lowest of any level – which barely features. Postgraduates are the most prepared, the most confident, the most likely to have clear career goals, and the least satisfied with the support they received. The insight report mentions the figure once and immediately pivots to the Level 4/5 satisfaction rate (76 per cent).

Given that postgraduate taught (PGT) students now make up roughly 29 per cent of the graduate population, and that many institutions are financially dependent on international PGT fee income, the decision not to interrogate this finding is striking.

In our forthcoming analysis, PGT and postgraduate research (PGR) students carry the largest positive residuals in the career confidence dataset – but this appears to be driven by their own maturity and goal clarity, not by anything the institution is doing for them. OfS’ own data is consistent with the reading, but it doesn’t draw the conclusion.

Central casting

The insight report does various things that go beyond summarising the IFF research, and the editorial choices matters because they reveal the regulator’s theory of change – or, more precisely, the theory of change it wants the sector to adopt.

The opening move is to contextualise away the preparedness finding. The insight report introduces progression rate data – 71.2 per cent for full-time first degree graduates into managerial/professional employment or further study – that doesn’t appear in the IFF research, and states that “the data suggests graduates do generally go on to successful careers, and their skills are in high demand.” This is deployed to reframe the 50 per cent preparedness figure as potentially a perception problem rather than an institutional failure.

GFI’s research on Uncovering skills is then cited to argue that students may not be able to “recognise or articulate the value of their experiences” – effectively suggesting that graduates are more prepared than they think they are, and the gap is one of confidence and self-awareness.

That’s a reasonable hypothesis, but OfS should be cautious about it. Our own data shows that career preparation confidence correlates at r=0.45 with feeling that life is worthwhile, and that nearly half of students in the lowest career confidence group score in the “low” category on the Office for National Statistics (ONS) wellbeing framework.

If the argument is that graduates’ perceptions of unpreparedness are inaccurate, you still need to reckon with the fact that those perceptions have real and substantial wellbeing consequences. Telling students they’re more prepared than they think doesn’t make them feel better – or, for that matter, more prepared.

The next move is to frame the solution primarily in terms of careers services. The insight report’s “points to consider” are structured around visibility, accessibility, tailoring, and coordination of careers and employability support.

This is where GFI’s influence is most visible – the framing assumes that the right answer is better-resourced, more visible, more tailored careers services, echoing Edmondson’s call for “appropriate institutional resourcing and prioritisation.” The case studies – City St George’s and UWE Bristol – both showcase sophisticated central institutional infrastructure, from personalised tracking systems to lifelong careers support and bespoke programmes for care leavers.

The difficulty is that the IFF’s own data doesn’t especially support this framing. The most common sources of institutional support were academic staff (46 per cent) and course teaching (41 per cent) – not the careers service (33 per cent). Informal support from individual staff members was rated more useful (56 per cent) than formal institutional support (50 per cent).

The dominant qualitative complaint wasn’t about the existence of careers services – it was about how generic they are:

There was nothing industry-specific, there was nothing course-specific. There was nothing person-specific. It was all just like generic CV tips.”

Our analysis will show something sharper – students in the lowest career confidence groups rarely mention careers services at all. When they do, the references are overwhelmingly negative. References to “my course,” “my lecturers,” and “my supervisor” outnumber references to “the careers service” by roughly ten to one, for both credit and blame.

The students who feel most disconnected from career preparation haven’t used the careers service and found it wanting – nobody on their programme has ever drawn a connection between their studies and their futures.

That’s an overall student experience design problem, not a careers service problem, and the insight report’s checklist of careers service improvements – however sensible each individual item – doesn’t address it.

Course correction

The insight report’s implicit theory of change runs something like this – graduates would feel more prepared if careers services were more visible, more accessible, more tailored, better coordinated with academic departments, and better resourced.

That’s fair enough, as far as it goes – each of those things would probably help at the margin. But it amounts to “do what you’re already doing, but better and more of it,” which is a fairly thin policy proposition from a regulator that holds Condition B3 as a central pillar of its framework.

What’s missing is any engagement with structural questions about where career preparation sits in the institutional architecture. The insight report asks institutions to consider “how can you embed employability support and information about steps to further study in course delivery?” but doesn’t distinguish between fundamentally different approaches.

You could make career connections visible within existing discipline teaching – the history lecturer explaining what analytical skills an essay develops. You could credit professional experience alongside academic teaching – placements, professional practice modules, and employer-briefed projects within the degree structure.

And then there’s more optional central provision – careers fairs, workshops, and drop-in sessions. They are structurally different interventions with different reach profiles, and lumping them together as “employability support” obscures the critical question of which models actually work.

Our data will suggest the answer is quite specific. Credited separate provision – placements, professional practice modules, credited internships – is the most commonly named institutional mechanism among students with the highest career confidence.

Optional central provision peaks among self-selecting students who are already relatively career-confident. And the emerging sector approach of “injecting” employability content into existing academic modules – asking discipline lecturers to retrofit skills exercises – solves a problem that students don’t appear to have.

Across our qualitative responses from students in the lowest two career confidence groups, none ask for a skills module or an employability module. They ask for placements, for guidance on where the degree leads, and for practical experience. They ask their degree to include more, not their lecturers to teach differently.

The HE in FE outperformance finding – which both the IFF report and OfS’s insight report note without really interrogating – points in the same direction. FE colleges delivering HE have fewer resources, smaller careers teams, and less institutional infrastructure than universities. What they have is course structures that are vocational by design, with integrated work experience, close employer relationships, and clear professional endpoints.

If OfS’ theory of change is “improve careers services,” the FE colleges are an awkward counterexample – they’re producing the best outcomes with the least infrastructure of the kind the insight report recommends building.

Two tribes

Both the IFF report and OfS’s insight report note the subject-level variation without distinguishing between two structurally different problems. Our forthcoming analysis makes this distinction directly – students on academic courses who report low career confidence are characterised by directionlessness, while students on vocational courses who report low career confidence have a different problem – they know what they should be training for, but feel the course isn’t delivering the practical experience it promised.

The insight report treats these as the same problem, recommending that institutions “expand integrated placements, industry guest sessions and alumni mentoring schemes across all disciplines, including non-STEM disciplines.” That might be the right instinct for academic courses, but it misses the point for vocational courses where the issue is theory-to-practice delivery failure.

At the same time, the insight report’s emphasis on visibility and accessibility of careers services is more relevant for the academic disciplines where students are directionless than for the vocational disciplines where students have a clear destination but feel their course isn’t getting them there.

To be fair, the IFF research report – as distinct from OfS’s gloss on it – contains a ton of useful data. The barriers analysis (financial challenges 45 per cent, lack of work experience 43 per cent, lack of professional network 40 per cent) gives institutions concrete things to work on.

The finding that graduates who were unemployed at the time of the survey were less likely to have received informal staff support (58 per cent versus 70 per cent) but more likely to have engaged with CV and application support (77 per cent versus 70 per cent) suggests that transactional support is reaching struggling graduates while relational support isn’t – a distinction that has practical implications for how services are structured.

The high-tariff paradox – more engagement with support, lower satisfaction, more likely to identify lack of a professional network as a barrier – deserves a look too.

The focus group testimony is also valuable, if limited by the sample size (18 participants). The nursing graduate describing friends quitting within three months of entering the profession, the law graduate unprepared not for the work but for the unemployment, the first-generation student who found the shift from school-style “hand-holding” to university’s “you have to seek it out” model a significant barrier – these are voices that should inform institutional practice.

The suggestion of peer-to-peer support forums established before graduation, so cohorts can support each other through the first months of transition, is a good idea that costs very little to implement.

Sadly, the “points for universities and colleges to consider” at the end of the insight report, while individually reasonable, read as a checklist rather than a strategy. There’s nothing here that a careers service professional wouldn’t already know.

The question OfS doesn’t ask – and which its regulatory framework arguably should push it to ask – is whether these things actually work, and specifically whether they reach the students who are furthest from career readiness. That requires a different kind of evidence than retrospective graduate perception surveys, and a different kind of regulatory posture than “points to consider.”

Points to consider?

The insight report carries no compliance implications, no connection to B3 monitoring, and no indication that OfS will follow up on any of its suggestions. The “Explorations” branding signals research-for-the-sake-of-it rather than research that will drive regulatory action.

Fleming’s quote – “we hope today’s research will be useful to universities and colleges” – is the language of encouragement, not enforcement.

The more interesting question is whether OfS’s framing of this issue will survive contact with stronger evidence. The IFF survey, at 1,671 graduates, is a reasonable sample but its retrospective design and online panel methodology carry well-documented limitations – not least the likelihood that graduates who engaged with institutional support are overrepresented among respondents. The 88 per cent figure for having received “any” institutional support may look different in a sample that better captures graduates who fell through the cracks entirely.

Wonkhe’s own analysis, drawing on responses to the career preparation question across nine waves of survey work, will be published shortly. It tells a structurally consistent story – subject discipline is the dominant factor, vocational courses produce higher confidence, the disability gap persists after controlling for subject – but it also tells a story OfS hasn’t engaged with.

Career preparation confidence declines as students progress through their degree, not increases. The wellbeing correlation is substantial and concerning. The qualitative evidence points overwhelmingly to course design and departmental ownership, not central services, as the mechanism that actually moves students up the career confidence spectrum. And the emerging sector consensus around “embedding employability” in discipline teaching appears to solve a problem that students aren’t asking to have solved.

Whether OfS, and GFI whose framing it has adopted, will engage with these findings remains to be seen. At a minimum, I’d encourage a reading of the IFF research report in full – the data is richer and more challenging than OfS’s gloss suggests – and the students’ own words, when you let them speak without editorial mediation, point toward a rather different set of conclusions about what would actually help.

But one other thing while I’m on. Even if providers got all of this right, there’s a bigger problem that neither the IFF research nor the insight report acknowledges.

The insight report opens by linking its findings to Skills England and the post-16 white paper, framing the transition problem as one of alignment – universities need to prepare students better for jobs that exist and are growing.

But as we’ve explored at length on the site, the UK’s shrinking graduate premium isn’t an inevitable consequence of mass higher education. It’s a peculiarly British problem, rooted in a labour market that doesn’t generate enough roles that genuinely require and reward higher-level skills.

Models of economic growth are fascinating on this. If you look, for example, at Scotland’s National Strategy for Economic Transformation, universities are really only mentioned as incubators and research centres. Their production of graduate human capital is absent.

It matters for how we read the IFF data. When 32 per cent of graduates feel unprepared to secure work in their desired field, that isn’t just an institutional shortcoming – it may be structurally true. There aren’t enough graduate-level roles in the sectors some degrees point toward, particularly outside London.

And OfS, as a market regulator, has essentially no levers on this side of the equation – it can’t influence how many graduates there are in which subjects, where they study, what their credit represents, or whether employers invest in utilising their skills. Nor does it have advice for universities on how to drive change in that space. It can encourage institutions to improve their careers services, which is exactly what it’s doing here. But if the jobs aren’t there, the most beautifully tailored careers service in the world can only do so much.

So we end up with three nested problems. Universities are locating career preparation in central services when a much wider look at the student experience is warranted. Little work is being that understands the role of graduates in growth unless they become a PGR student or they enter Dragon’s Den. And so the system treats graduate transition as a supply-side problem that can be solved by preparing students better, when the demand side of the equation is what needs intervention.

Students, to their credit, seem to understand all three. The system keeps mishearing them.