What Makes a Graduate Worth Hiring Now?
Premier League clubs are owned by Russian oligarchs, Saudi sovereign wealth funds, American private equity consortiums, and the occasional petrostate vanity project. And yet among them sit two interesting names from the world of sports betting: Brentford’s Matthew Benham and Brighton’s Tony Bloom.
Benham is a former derivatives trader turned professional gambler. Not the romantic kind. The kind who builds statistical models and compounds small edges over thousands of events. He made enough money to buy his boyhood club, and when he took over Brentford in 2012, they were in England’s third tier. He brought trading-floor instincts to West London: searching for which metrics actually predict player performance, which signings represent mispriced value, and where the club is spending money out of habit rather than evidence.
That search led to what many saw as sacrilege in 2016: shutting down Brentford’s youth academy. The logic was brutal but difficult to dismiss. The club was spending around £2 million a year developing young players, only to see almost none make the first team. The best prospects were regularly snapped up by more prestigious London rivals. So Benham replaced the academy with a B-team model, signing hungrier players between seventeen and twenty-one who had been released or overlooked elsewhere, testing them in tougher competitive environments, and promoting the ones who proved they could actually play. It was a cold-blooded decision. It also helped Brentford reach the Premier League, before they later reintroduced a youth setup under top-flight pressures and standards.
If that leaves a sour taste, that is part of the point. Benham was asking a question many organisations hate asking plainly: is this system actually producing enough value for what it costs?
Strip away the football and sentiment, and plenty of employers may soon find themselves facing a similar question about graduate programmes. For the money, management time and development effort involved, are they generating enough return? Are the strongest graduates sticking around long enough to justify the investment, or are they simply being polished for someone else to hire? In a tighter economy, those questions get difficult. In a world where AI may automate or compress parts of entry-level work, they get more difficult still.
That does not mean the answer is to give up on graduates. It may mean the opposite. The organisations that keep investing in young talent will just need to get much sharper about what they are really hiring for, and how they assess it. Because if AI changes the shape of work even moderately, graduates and young professionals are likely to be among the first groups affected. They may also be among the best placed to help organisations adapt, especially where fresh technical instincts and lower attachment to old workflows become advantages.
So rather than argue about whether graduate programmes are dead, it is probably more useful to ask a narrower question: what skills are actually becoming more valuable, and how should organisations test for them?
Our take on graduate skills and assessment
Crudely put, most hiring managers are looking for two broad things: attitude and ability. I think that framing is still useful, but only if we sharpen it. Both words are too often used as vague placeholders. So below is a more practical version of what they can mean, and how organisations might actually assess them.
1. Attitude, or “the right attitude”
Seldom said but often thought, this is one of the biggest questions in the mind of any hiring manager. “The right attitude” is not a useless phrase, but it is usually too soft and too vague. In practice, it often collapses into shorthand for culture fit: pleasant, eager, easy to manage. Those things are helpful, but they are not enough.
A better framing is a candidate’s inner engine. More specifically, whether they show the kind of high-agency behaviour that matters in fast-changing environments.
I think of that in terms of three A’s: action, accountability and adaptability.
Action. Agency starts with a willingness for action. Strong candidates usually have evidence that they have made things, solved problems, or pushed beyond what was strictly required. With AI and modern tools lowering the barrier to entry, that evidence can take many forms: GitHub repos, web apps, blogs, experiments, small businesses, community projects.
Accountability. Action without consistency can just be enthusiasm with good PR. Accountability is about whether someone can be relied on over time. Look for signs that they can carry long team projects through, handle responsibility, and contribute to something larger than themselves.
Adaptability. Some candidates can perform well in one narrow lane and then disappear the moment the terrain changes. Adaptability is what keeps someone useful when the brief is unclear, or the environment is unfamiliar. Resilience, resourcefulness and creativity all live here. So does something else that matters more than many employers admit: range.
South Africa is still a signalling culture. Many people are understandably hesitant to step outside their lane after university unless there is a certificate, course or formal badge attached to it. But genuine curiosity often shows up elsewhere: hobbies, side projects, deep interests, self-directed learning, reading across domains. Signals of polymathic tendency can be especially revealing of agency here.
2. Ability
This usually refers to a candidate’s hard skills, often the technical competencies acquired through formal education. Most companies already know how to assess those reasonably well. The more overlooked question is whether a candidate has the auxiliary skills needed to turn technical competence into organisational value, especially in environments where AI is entering the workflow.
A few stand out.
Systems thinking. It is no coincidence that many early machine learning and AI systems that actually made it into production were built by people who could think beyond the model itself. A brilliant piece of analysis means little if it cannot fit into the organisation’s infrastructure, data flows, incentives and workflows. Even where candidates do not arrive with this instinct fully formed, it is worth identifying early because it compounds fast.
How to test it: Give candidates a broken or inefficient process inside a fictional organisation and ask them to map the dependencies, bottlenecks and knock-on effects before proposing a fix. The key signal is whether they zoom out before zooming in.
Design thinking. If systems thinking is about how a solution fits into an organisation, design thinking is about whether the solution fits the human beings expected to use it. It speaks to choosing the right problem, defining the right success metrics, thinking about usability, and recognising when something technically correct is still practically dead on arrival.
How to test it: Show candidates a working but badly designed internal tool, report or dashboard and ask how they would improve it. Strong candidates usually start by asking who the user is and what the tool is meant to achieve.
Critical thinking. This is the quality of reasoning itself. Can the candidate analyse unfamiliar material, form hypotheses, test assumptions, and arrive at conclusions that are defensible rather than just confident? This becomes even more important in an AI-heavy environment, where polished nonsense is often cheaper than ever to produce.
How to test it: Give candidates a short unfamiliar dataset, case study or brief with conflicting signals. Watch for whether they state assumptions, flag uncertainty, and resist forcing a tidy answer where the evidence does not support one.
Graduate talent is not going away. But the shape of entry-level value is changing, and organisations will need to be more deliberate about what they look for.
At SimplyfAI, we have been thinking about this problem for a while. We are currently testing an interactive assessment tool designed to help measure some of these underlying skills in a more practical, decision-useful way. More on that soon.
