Where AI Belongs in a Recruiting Team That Got Smaller

Suppose a mid-market company had 14 recruiters before the 2023 hiring freeze and has nine today. Hiring demand has come back since then. The recruiters haven't.

You could describe that as a team a third smaller, and it is, but a more useful way to see it is this. The company still wants roughly the same number of hires a year, and it now has fewer people to make them. If those nine recruiters work the way the fourteen did, there's only one way it can go. Roles take longer to fill, candidates fall out halfway through the process, and hiring managers end up doing coordination work that was never supposed to be theirs.

When companies made cuts like this, the assumption in the room was usually that AI would pick up the slack. Software would handle the repetitive work, a smaller team would focus on the parts that need judgment, and costs would come down. I don't think that reasoning was wrong in principle. Where it went wrong was about which part of the process AI would take over, and when.

Most of the teams that got cut weren't running much AI at all. They had an ordinary ATS, LinkedIn Recruiter for sourcing, and people scheduling interviews by hand. When the headcount went, those manual steps didn't get automated. They got handed to fewer people, who are now at capacity on a volume the old team could barely manage.

A wider mouth on the same funnel

And the companies that did spend on AI after cutting mostly spent it at the top of the funnel: programmatic job advertising, automated LinkedIn outreach sequences, job descriptions written by a model. All of those produce more applicants. None of them reduce the human work of screening those applicants, qualifying them, and moving the good ones forward.

So you get a funnel with a wider mouth and the same narrow middle. More people apply, the same few recruiters are there to respond, and responses get slower. Candidates wait. Some of them take another job while they're waiting.

There's an odd consequence to this. A company that invested heavily in advertising can end up filling roles more slowly than one that didn't, because it raised volume without raising the capacity to deal with it. I suspect some teams are in exactly that position and can't work out why the extra spend made things worse. The shortlist gets harder to build as the pile of unscreened candidates grows, and nobody new has been hired to work through it.

The step worth taking out of human hands

If you're going to automate one step, I'd pick first contact: the structured qualification conversation that decides whether a candidate is worth a recruiter's time. Relationship building, hiring manager conversations, offer negotiation, and the moments that make a candidate commit should stay with people. First contact is where the volume is. It needs to happen whenever people apply, at whatever volume arrives, and produce the same structured result whether ten people applied or four hundred.

To put rough numbers on it, a team that could manage 300 qualification calls a month can, with software doing those calls, run a campaign that brings in 1,200 applicants and still come out with a usable shortlist. The recruiters do about as much work as before, only on a group that has already been filtered, and they spend their time where their judgment counts.

First contact is the step we built Asendia to do, and we made it a phone call on purpose. A qualification form can check boxes, but it can't notice that an answer was vague and ask about it. A spoken conversation can, and the follow-up is often where you learn whether someone is worth a recruiter's time. So when someone applies, Asendia phones them, often that same evening or over the weekend, and runs a structured qualification interview against criteria set for that particular role, adapting its questions to what the candidate says. Instead of starting the morning with a stack of unreviewed applications, the recruiter starts with a ranked list of qualified candidates and excerpts from each conversation, already in the ATS. Screened candidates show up in the same workflow as before, only sooner and already spoken to.

For an in-house team running lean, that closes much of the gap between what the reduced team can do and what the business needs from it. The limit on hiring stops being how many people you employ and becomes how fast work moves through the pipeline, which is a much easier problem to work on.

If you're trying to tell whether changes like this are working, I wrote about which numbers reflect real pipeline health and which ones give false comfort in the post on recruitment KPIs in a post-AI world.

The bet companies made in 2023 and 2024, that AI would cover for smaller teams, was early more than it was wrong. The work that got cut was mostly first-contact work, which is high-volume, time-sensitive, and structured, and that's the kind of work software is good at. Many of those companies simply never automated it. Before rehiring, it's worth looking at where candidates in your pipeline actually sit and wait. My guess is that for a lot of teams it's the first conversation.

Ready to transform your hiring strategy? Schedule a Demo with our founders today!