Writing

When every CV looks good, where does the signal come from?

Josh del Rio · 2 November 2026

Updated October 2026.

Hiring has always started with a CV. In the last three years the volume of CVs has more than doubled, and AI has made most of them polished, tailored and hard to tell apart. The CV still matters: what someone has done is relevant to what they’ll do next. But as the only filter at the top of the funnel, it now tells you less than it used to. In our own data, covering more than 17,000 candidates screened since late 2025, 43% of the people we hired weren’t in the top quarter of applicants on their CV. Finding the right people now depends on adding information that’s harder to produce on demand.

What’s changed since AI arrived?

More applications, fewer interviews per application, and more time spent sorting.

Greenhouse, whose data covers more than 6,000 companies, reports 244 applications per job in 2025, up 111% since 2022. Ashby’s data shows applications per hire roughly tripled, from about 100 in early 2021 to about 291 in early 2026. Over the same period, the share of applications that lead to an interview roughly halved, to between 3.6% and 4.7%. In Australia, SEEK reported applications per job ad at record highs this year.

Most of that growth comes from how cheap applying has become. Greenhouse found around three-quarters of candidates now use AI in their job search, and nearly half say they use it specifically to get past automated filters.

Why did the CV work as a first filter?

Because writing a good one took effort.

As our founder Josh del Rio, who spent years in systematic trading, puts it: “A signal is only useful while it’s costly to fake.”

A good CV used to be a reasonable signal for exactly that reason. It took time to describe your experience clearly and fit it to a role. That effort tracked, imperfectly, with ability and with genuine interest in the job. That’s why businesses could use the CV to decide who to look at more closely.

What has AI done to the CV?

It has lowered the cost of a polished, tailored CV, so polish and tailoring tell you less than they used to.

Hiring managers see this directly. In Resume Genius’s 2026 survey, 76% said the overall quality of CVs has improved in recent years, and 76% said AI-written CVs make it harder to understand what a candidate actually did. In Robert Half’s survey of 2,000 hiring managers, 65% said AI-generated applications have made hiring harder, and 58% said truly qualified candidates are harder to identify than a year ago.

Automated screening adds a further twist. Researchers at the University of Maryland, the National University of Singapore and Ohio State found that AI models used to screen CVs consistently favour CVs written by the same kind of model, even when the underlying quality is the same. In simulated hiring across 24 occupations, candidates whose CV matched the screening model were 23% to 60% more likely to be shortlisted than equally qualified candidates. The biggest gaps were in sales and accounting.

Clarity itself isn’t the problem. A large field experiment published in Management Science found that helping jobseekers write clearer CVs increased hiring and wages, with no drop in employer satisfaction, because clearer writing helps employers see ability. The issue is narrower: when the content of the application becomes cheap to produce, it separates people less well. The clearest measurement of that effect comes from an adjacent setting. On a large freelancing platform, after an AI writing tool launched, the link between how well an application fitted the job and whether it got a callback fell by about half.

Why doesn’t screening harder fix it?

Because both sides are now automating against the same document.

Candidates use AI to apply. Employers use AI to cope with the volume. Candidates adjust again. Filtering the same document more aggressively, or trying to detect AI writing, keeps the business inside that loop. The answer isn’t more filtering of the CV. It’s adding information the CV can’t carry.

What does our own data show?

That the CV is a useful starting point and a weak gate.

We looked at the roles we’ve filled since October 2025, when our current process was in place across every campaign. That’s more than 17,000 candidates screened. For each person we hired, we compared their CV rating with everyone else who applied for the same role.

What found those people was the information we gathered after the CV. The hires outside the top quarter on CV typically sat in the top third of their campaign on attitude. And the disagreement runs the other way too. Of the candidates with the strongest CVs who also completed our assessments, 43% sat below the middle of the field on attitude.

Where does better signal come from?

From information that’s hard to produce on demand, and that’s linked to what actually predicts performance in the role. Two things stand out.

1. Validated assessments, linked to your top performers

The most useful addition at the top of the funnel is a consistent measure of how someone works and what drives them, compared against the people who already succeed in the role.

In practice, that means:

It holds up because it measures what a polished document can’t show, under the same conditions for everyone. It needs care: evidence that the assessment predicts performance, a short and respectful candidate experience, and a design that can’t simply be handed to an AI. Our guide to hiring well for the roles you hire most covers how to build the profile.

2. Going to the people who aren’t applying

The best people for a role are often not applying for anything. Reaching them takes more than posting an ad and screening what arrives:

It holds up because a track record and a reputation in the market are much harder to fake than an application, and because someone approached for a reason starts from a stronger position.

How the two work together

Sourcing widens who you see. Assessment tells you who fits. The CV doesn’t disappear: experience still matters, and it’s still part of the picture. It just stops being the gate that decides who you look at.

How we approach it

For every role we run:

Hiring for a role now? Tell us about it, and we’ll tell you what the market looks like for it and when you could expect a shortlist, before you commit to anything.

Josh del Rio is the founder and CEO of Ploomo, which runs hiring for growing businesses so every role is built on evidence, not instinct. He has coached hundreds of leaders and founders.

← All writing

Start here

See what it looks like for your business.

Pick a live role, or one you’ve been trying to fill. In 30 minutes we’ll show you how we’d approach it and what you’d see at each stage. If it’s not a fit, we’ll say so.

For anything else, reach us at contact@ploomo.io