Early careers recruitment has a volume problem.

That’s how it’s usually framed, anyway. More applications than ever. Bigger funnels. Greater reach. On the surface, it looks like success. The numbers go up, the dashboards look healthy, and somewhere along the line “more” became shorthand for “better”.

But spend five minutes inside that volume and the picture changes.

When hundreds of candidates apply for a single role, decisions don’t get easier. They get harder. What looks like abundance from a distance starts to feel like interference up close. The signal weakens. Patterns blur. Judgement slows down. You’re not choosing from a richer pool; you’re wading through a noisier one.

And yet, we still treat volume as a win.

That’s the first mistake.

Because what’s happening in early careers recruitment isn’t a temporary spike. It’s something more structural. Application numbers have been rising faster than vacancies for years now, and in some parts of the market they’ve tipped into the high hundreds per role. Not because every employer suddenly became more attractive, but because candidate behaviour has changed.

Applying has become easy. Deciding has not.

The friction that once forced a bit of reflection has largely disappeared. You can apply to a role in less time than it takes to properly understand it. Autofill has replaced effort. One-click has replaced intent. And where there used to be a small moment of hesitation, is this actually right for me?, there’s now very little to interrupt the impulse to apply.

Layer generative AI on top of that and the shift becomes more pronounced. Candidates aren’t just applying more; they’re applying more convincingly. Answers are polished, aligned, and tailored at a speed no human recruiter could realistically keep up with. The written application, once a rough proxy for motivation or fit, is increasingly a proxy for how well someone can instruct a machine.

None of this is irrational from the candidate side. Quite the opposite.

Confidence among early talent is fragile. Many are as worried about failing in a role as they are about not getting one. In that context, applying widely feels like the sensible thing to do. If you don’t know which doors will open, you knock on as many as possible. The system rewards that behaviour, so it repeats.

What you end up with is a market full of activity, but not necessarily intention.

And that’s where the problem starts to bite.

There’s a comforting belief in recruitment that more applications must mean more choice. More diversity of thought. A better chance of finding the right person. It sounds logical, but it quietly breaks down at scale. Because while applications can grow almost infinitely, the capacity to assess them can’t. Recruiter time, hiring manager attention, interview quality, these things don’t expand just because the top of the funnel has.

So the marginal value of each additional application drops, while the cost of processing it stays stubbornly real.

At the same time, the quality of the signal begins to erode. If large numbers of candidates are applying with limited understanding of the role, and presenting themselves through increasingly standardised, AI-assisted narratives, then each application tells you less than it used to. Not more.

You start making decisions in a fog.

And fog is risky. It increases the chances of hiring the wrong person, but also of missing the right one entirely, the candidate who didn’t optimise their application quite as effectively, or who got lost somewhere between filters and fatigue.

Faced with that, most organisations respond in a predictable way. They don’t question the volume. They build around it.

More screening questions. More assessments. More automation. Longer processes that promise rigour but often deliver distance. It feels like control. In reality, it’s adaptation. The system bends to accommodate the noise rather than asking where the noise is coming from.

Candidates, unsurprisingly, adapt too. If the process feels impersonal, they hedge their bets. If it feels optimised, they optimise back. More applications, more AI, more attempts to beat a system they don’t entirely trust.

And so the cycle reinforces itself.

More volume leads to more filtering. More filtering leads to more gaming. More gaming leads to more volume.

At no point does anyone stop to ask whether the starting assumption, that more applications are inherently good, was ever true in the first place.

Because underneath all of this is a simpler issue.

Conviction.

A significant proportion of candidates are applying without a clear sense of why this role, in this organisation, is right for them. Not because they’re careless, but because the system doesn’t really help them figure that out before they apply. It encourages access and speed, not understanding.

From their perspective, applying broadly is rational.

From yours, it creates noise.

You see a full inbox, but very little evidence of genuine intent. And intent is the thing that actually predicts what happens next, who accepts, who performs, who stays.

Which raises a different question.

Not “how many applications did we get?”

But “how many people truly understood what we were offering, and chose it anyway?”

That’s a much harder metric to chase. It doesn’t spike overnight. It can’t be bought through media spend or engineered through minor tweaks to a job description.

It requires a shift in where you focus.

Upstream, rather than downstream.

Clearer, more honest communication about the reality of the work. Not just what’s attractive about it, but what’s demanding. A sharper articulation of who tends to thrive, and who might not. Content that helps candidates make a decision, rather than simply encouraging them to apply.

In other words, designing a system that filters before the application, not after it.

That might sound counterintuitive in a culture that still celebrates big numbers. But reducing the wrong applications is one of the most effective ways to improve the right ones.

This is also where most organisations lack visibility. They can see volume, but they can’t see intent. They know how many people applied, but not how many understood. They measure outputs without really understanding the behaviours driving them.

That’s the gap tools like the Future Talent Barometer are designed to fill. Not by telling you whether your campaign “worked”, but by showing how your audience is thinking and behaving over time. Confidence, trust, use of AI, willingness to engage, signals that sit beneath the surface of your funnel, shaping what eventually lands in it.

Because a rise in applications alongside a drop in confidence isn’t success. It’s a warning sign.

Equally, a smaller, more stable pipeline paired with stronger understanding and intent might be exactly what progress looks like.

Seen through that lens, application inflation stops being a resourcing headache and starts to look like something more useful: feedback. A signal that the system is currently optimised for speed and access, rather than clarity and commitment.

Change the system, and the behaviour changes with it.

The outcome is unlikely to be more applications.

It’s likely to be fewer. But better.

And in a market where everyone is competing for attention, but very few are designing for understanding, that shift, from volume to intent, isn’t just operationally helpful.

It’s a genuine advantage.