Every recruitment leader I talk to says they need more candidates.

They don't.

I'd say 80% of the time, candidate volume is fine. The real problem lives elsewhere. But "we need more candidates" is the most comfortable diagnosis because it points outward — at the market, at sourcing tools, at LinkedIn's algorithm, at anything that isn't us. The actual bottleneck almost always lives inside the team, and that's a harder conversation.

So. Where's it actually living?

The hiring manager doesn't know what they want.

The brief was vague. The job description is a copy-paste from a similar role two years ago. The first three candidates get rejected for reasons the manager can't articulate beyond "not quite what I had in mind." The recruiter goes back to sourcing thinking the problem is candidate quality. It isn't. The problem is that the hiring manager hasn't yet figured out, in concrete terms, what good looks like — and they're using interviews to figure it out, candidate by candidate.

Fix: a 45-minute intake meeting at the start of every role. Three questions. What does this person do in their first 90 days? What's the failure mode you're worried about? Show me a CV that would be a "yes" and one that would be a "maybe". If the manager can't answer those clearly, the role isn't ready to open.

Interviewers aren't calibrated.

Five people interview the same candidate. They come back with five different reads. The decision goes the way of whoever spoke loudest in the debrief, or whoever the manager trusts most that week.

Fix: scorecards, structured questions, written feedback before the debrief. None of this is new. Most teams know they should do it. Most teams don't, because it feels like overhead, and they don't see the cost of not doing it because the cost lives in bad hires nine months later.

Decisions take three weeks.

Final interview Thursday. Debrief scheduled for "next week sometime." Reference checks haven't been initiated. Comp hasn't been signed off. By the time the offer goes out, eighteen days have passed and the candidate has accepted somewhere else, or worse, they've accepted yours and are already in a slow drift toward declining because the momentum is gone.

Fix: a hard 5-business-day SLA from final interview to offer. Everything that needs to happen — references, comp approval, draft offer letter — gets pre-staged. Make it a rule. Enforce it ruthlessly. If you can't hit it consistently, your process has more bureaucracy than the role can afford.

Nobody's tracking acceptance rate by source.

Some sources convert at 85%. Others at 40%. If you're not tracking it, you're spending the same effort on both. You're also using the wrong sources to project pipeline capacity, which leads to thinking you need more candidates when you actually need different ones.

Fix: tag your candidate sources cleanly in the ATS. Pull the acceptance rate every month. Reallocate.

You'll notice none of this involves more candidates. None of it involves AI either. Mostly it involves discipline that's annoying to put in place and unglamorous to maintain.

But here's what happens if you actually do it: the same candidate volume produces dramatically more hires. The recruiters stop burning out chasing pipeline that gets wasted at the bottom of the funnel. The hiring managers learn faster because the structure forces them to. And the AI investments you make later actually compound, because you're putting them on top of a reliable operation instead of a broken one.

I have never — never — worked with a team where "more candidates" was the actual answer.

I've worked with maybe one or two where it was even part of the answer.

If you're convinced your bottleneck is sourcing, the most useful thing you can do is spend a week assuming the opposite and seeing what you find.