Why Your Job Ads Aren't Attracting Candidates

A note on this rewrite

This article was originally published in November 2021 under the title "The Great Resignation." At the time, employers across Australia were openly struggling to attract candidates. Resignation rates were high, job ads outnumbered applicants in many sectors, and HR commentary was full of articles about "what the Great Resignation means for employers."

I argued then that a lot of the candidate-attraction problem was self-inflicted: employers were writing bloated, obfuscated, internally contradictory job ads that were filtering out exactly the candidates they wanted to attract. The Great Resignation was the moment that problem became visible — but the problem itself wasn't caused by the Great Resignation. It existed before, and it has continued.

Five years on, the labour market has rebalanced. Unemployment in Australia sat at 4.3% in March 2026, up modestly from its 2024-25 lows but still relatively low historically. Job ad volumes have fallen year-over-year. Application volumes per role have surged. Many corporate and APS roles now receive 200-500+ applications, and recruiters are openly drowning. The market has shifted from candidate-driven to something closer to balanced — in some sectors, employer-driven.

The Great Resignation framing is now historical. But the underlying critique — that employers are still writing job ads that filter out the people they want to hire, and that the application process is treated as a screening proxy for things that should be screened at interview — is more relevant in 2026 than it was in 2021. So I've kept the URL, kept the bones of the argument, replaced the dated framing, and added what's changed.

What's actually changed since 2021

Four things have shifted that materially affect the candidate-attraction picture:

• The labour market has rebalanced. Unemployment is up moderately from its 2022-23 lows. Job ads have declined. Candidate competition per role is higher. The acute "we can't find anyone" panic of 2021-22 has faded for most sectors — though it persists in healthcare, aged care, trades, and some regional and specialist niches.
• Application volumes have exploded. A typical mid-level corporate or APS role now receives several hundred applications. The bottleneck has flipped from finding candidates to finding the right candidates inside the pile. That makes the quality of the screening process the limiting factor, not the volume of the candidate pool.
• AI is on both sides of the application. Candidates are using AI to draft applications. Employers are using AI (or automated keyword-matching) to screen them. Around 49% of AI-generated resumes are now dismissed in the first round of screening, according to a 2025 study of 3,000 hiring managers. We've written more about that in our guide on editing AI-generated job applications. The arms race makes the obfuscation problem worse — bad PDs trigger bad applications, both AI-generated.
• The signal-to-noise problem is now structural. When a role gets 400 applications, the recruiter's challenge isn't "is this candidate suitable" — it's "can I get from 400 to 20 in less than two hours." If your position description doesn't clearly tell candidates what to put forward, the recruiter can't tell who's relevant either.

The thread that connects all four: clarity in the job ad has become the single highest-leverage hiring practice an employer has. Not advertising spend. Not employer brand. Clarity. The next 800 words are about why most employers are bad at it — and what to do instead.

The real reason employers can't fill roles

The reason our business exists is because of deliberately designed barriers to entry. Selection criteria, formatted PDs, structured pitches — these were originally designed to stop the wrong person from getting the job. Over time they've morphed. Now they often stop the right people, too.

The pattern looks like this. A senior recruiter or hiring manager sits down to write a position description for a role they need to fill. They have ten minutes. They reach for the last similar PD the agency or company ran, copy it, change a few details, add anything they're worried about leaving out. The PD ends up at three pages with thirteen duties and sixteen selection criteria, of which most overlap.

Then the PD goes live. The application closes a fortnight later. The recruiter is now looking at 380 applications. They have one afternoon to get to a shortlist of fifteen. They keyword-match against their own bloated criteria, dismiss anyone who didn't address all sixteen, and hand the survivors to the panel. The candidates who would have been excellent at the job — but who didn't have time to address sixteen items, or who recognised that some of them weren't really capabilities, or who simply didn't apply because the ad signalled an unserious organisation — never made it into the shortlist.

Then we hear: "we couldn't find anyone good."

Worked example: 16 criteria → 4

The example we used in the original 2021 piece still holds up. A federal regulatory and land management role, with a salary around the Australian average — at the time, around $90,000; in 2026, the average full-time salary is roughly $106,000 according to the most recent ABS data, but the analysis is unchanged. The right person for this role was defined using a three-page position description with the following profile:

Of the ten "essential" criteria in the original ad, the genuinely distinct capabilities boiled down to four:

Plus one item that wasn't a capability at all but a hard requirement — a current driver's licence and the ability to drive and maintain a 4WD over long distances. That belongs on the application form as a yes/no field, not as a written selection criterion.

If you've already written thirteen detailed domains of work, why repeat them in the criteria? The repetition obscures the point. If the candidate has read the duties and the four core capabilities, they can put together a good application. If they haven't, their application will be bad and you can rule them out on its merits. Only people with some relevant experience and interest are clicking on your ad in the first place. Why make it hard for those people?

Worked example: repetitive teaching criteria

The second example from the original piece — a state-government teaching role with sixteen criteria — was an even worse offender. Each of its ten criteria required at least half a page of response. But almost all of them asked essentially the same thing. Five pages of repetition, and no recruiter is reading five pages and remaining excited about the candidate at the end.

Reduced honestly, the role had three real capabilities:

Everything beyond those three was repetitive or self-evident. How five pages of selection criteria repeating the same content helps with assessment is beyond me — and, I suspect, beyond most of the recruiters who have to read them.

The "internal policy" criterion problem

One criterion in the same teaching role deserves particular attention because the pattern still appears regularly across federal and state government PDs in 2026:

Government entities should not be excluding people from outside the public sector. But this criterion refers to an internal policy — one the ad doesn't link to. Maybe a candidate could find it. But even if they did, an external candidate cannot possibly know how this policy has been operationalised inside the Department. They have no examples of times they've complied with a policy they've never had to apply.

How is a non-Departmental candidate meant to compete on this criterion? All they can honestly say is "I've read it." HR should be making candidates read it as part of induction anyway. As a selection criterion, it does nothing except favour internal candidates over external ones — without saying so openly.

The honest version of this criterion is either "internal candidates only" (a different decision, with different consequences), or "commitment to Aboriginal education and the cultural capability to operationalise policy in this area" — which is a real capability, and which an external candidate can evidence.

AI on both sides of the application

This wasn't a factor in 2021. It is now. AI has changed the dynamics on both sides of the hiring process, and both changes amplify the bad-PD problem rather than fix it.

On the candidate side: a large share of applications are now AI-drafted, fully or partly. The patterns AI generates — stock language, hedged claims, grammatically perfect but emotionally flat narrative — are exactly what bad selection criteria invite. If your ad is sixteen criteria of overlapping corporate-speak, an AI draft is the rational candidate response. The applications you receive are then thin, generic and indistinguishable from each other. You blame AI. The cause was upstream.

On the employer side: a growing number of organisations now use AI or automated keyword-matching to screen the application pile down to a manageable shortlist. AI screening rewards exact-keyword matches and penalises candidates who reframed your criteria into more direct language. Combined with AI-drafted applications on the other side, the result is a closed loop in which two AIs negotiate the shortlist while the human candidates and human hiring manager are largely cut out of the process.

The fix isn't more sophisticated AI. The fix is upstream: write a job ad that doesn't require AI to navigate. Four real criteria, written in plain language, mapped to actual deliverables. AI-screening then rewards candidates who actually have those capabilities, because those are the candidates who can write about them clearly.

Five questions before you post

Before sending the next job ad to the recruiter or pressing publish, ask yourself five questions. They take ten minutes. They will materially improve the quality of the candidates who apply.

And one final test that takes thirty seconds: think about the last batch of written applications you read. How long did you actually spend reading each one? If your honest answer is "less than two minutes," then asking candidates to write five pages of selection criteria response was always a waste of everyone's time.

A simple PD checklist

For the next role you write, run the position description through this checklist before publishing.

A closing note

The Great Resignation as a media story has ended. The labour market has rebalanced. Application volumes are higher than they've been in years. None of that has made the underlying problem go away — it has made it harder to see, because you're now drowning in applications instead of struggling to attract them.

The hiring problem is still, fundamentally, a clarity problem. The candidates you want — the ones who are good at their current role, well-supported, and selectively considering moves — read your job ad in the same way they read everything else. Carefully, briefly, and with a low tolerance for nonsense. If your ad makes them work to figure out what you actually want, the rational move is to apply somewhere else.

Make the ad clear. Make the response process proportionate. Trust your interview, your reference checks, and your probation period to do their actual jobs. The candidates will find you.

Read more

---