Faster at the wrong thing

A mate told me a story over the weekend that I’ve heard a thousand times. A recruiter messaged him a couple of weeks ago about a role he “would be perfect for”. He was interested, went back to them, and got ghosted, writes Fraser Rix.

He asked me why that happens.

Honestly, us recruiters have a lot of calls to get through and a lot on, so I get it. I’ve been guilty of it myself. But it got me thinking about what happens to that problem when you automate it and scale it, because that’s what the whole industry has been doing for the last couple of years.

I’m not slamming the tools; they are awesome! I use them every day, and I’m thankful for every second they give me back. AI will write the ad, post it, find the people, rank the applications, book the interviews, and take the notes, and it does most of that well.

But have a look at that list. Each one is about doing something faster, cheaper, or at scale. None of them decides whether the hire actually works out. And everyone bought the same tools around the same time, so any head start is gone. Your competitor’s ad went live just as fast as yours did.

The candidates have the same tools, too. Michael Page’s 2026 Talent Trends report, which is global and covers about 60,000 people, found that two in three job seekers had used AI on their CVs, and recruiters told the ABC back in May that applications now come in at scale and all read the same. So, you’ve got an ad written by AI, pulling in applications written by AI, screened by AI. More noise, less signal, and nobody in that chain has made a decision yet.

Here’s where I see it going wrong.

Outreach first, because that’s my mate’s story. A recruiter can now send a personal-sounding message to 200 people before lunch. The problem is, it worked. Forty people say yes; there’s one recruiter, and so the business becomes the one that approached someone, got a yes, and... went quiet. On a passive candidate, who’s worth more than any applicant you’ll see this year. Nobody needed 200 messages. They needed the right 20, from someone who’d answer.

Screening’s the one thing everyone’s excited about, and sometimes it’s exactly what you need. High-volume roles with objective screening criteria: you can cut 600 applications to 100 and rank them before you’ve finished your morning coffee. But in a lot of cases, we don’t have objective screening criteria; our needs are more subjective and harder to rank, and what are we ranking them against? I’ve lost count of the briefs I’ve been handed that were a job title, a salary band, and a list of must-haves someone lifted off the last ad. That won’t get you better results!

Assessments and psychometric tests are things I’ve used a lot, and they can add a lot of value, particularly in making the process more inclusive for participants. Set up correctly and pointed at the right thing, they can be so useful and create a really engaging process, even if it’s automated. Set up incorrectly, generic, not really assessing the right things, and the signal is weak, and the process doesn’t resonate or produce the results you need. I’ve shaped them differently over the years, too: focused on key elements of the role, gone micro, gone macro, left it as a video interview with one question. The best results I’ve found have always come from a day in the life, something the person will actually have to do in the role, so that they can see the role and you can assess them in their best light. Otherwise, just do the interview.

Scheduling, I’ll admit, is a genuine win and a game changer. Anyone who’s chased three interviewers across three diaries for five candidates knows how many hours that takes. But it was never what held the hire up. The thing holding the hire up is a hiring manager who won’t prioritise interviews, turns up without reading the CV, or takes a week to give feedback. The tool will find the slot. It can’t make anyone decide.

The tools will all promise better matching, less bias, more consistency, and better assessments. And fair enough, totally worth testing to see if they are right for your situation. But a matching tool learns from the hiring decisions you’ve already made, so mostly what you get is the same decisions, faster and with more confidence behind them.

So where’s the value now? It’s at the front, upstream, before the process gets agreed. It’s never been more important to know exactly what the role is, to have a hiring manager who can tell you what good looks like, and to have a pitch to the market that’s true.

And it’s at the back, after the tools have done their bit, the follow-through. Working out whether this person will actually work in your business, persuading them, closing, and aligning the role to their ambitions.

If your process was good to start with, the tools make it very good. If it wasn’t, you’ve bought a faster way to make the same mistake, and more of them.

The tools have made all the numbers look better. Still, when you speak to people who are going through these processes, they aren’t getting the value from the experience that they used to, and that’s where the opportunity sits: to be more authentic, more selective about the parts that you automate and the parts that you don’t, and to stand out against the noise.

My mate’s recruiter wasn’t rude. They were efficient. The message went out faster than ever, but nobody got any quicker at replying. The whole industry is racing to automate the sending, so when are we going to start automating the accountability?

Fraser Rix is the founder of Calder Talent.

Want to see more stories from trusted news sources?
Make HR Leader a preferred news source on Google.