Field notes · July 29, 2026 · 4 min read
How Recruiters Can Prioritize Existing Candidates Before Sourcing New Ones
> Editorial Note (August 2026): This post was written during an earlier phase of Voco, when our exploration focused heavily on building software for recruiting teams. While the observations about friction in those workflows remain true, Voco now operates more broadly to build custom AI systems for operational teams. You can read more about why we stopped starting with software.
A recruiter opens a new req with, realistically, hundreds or thousands of candidates already sitting in their database from past searches. The problem was never a shortage of people. The problem is that "search the database" returns a list, not a ranking — and a recruiter staring down a list of forty plausible matches with no way to sort them tends to default to the thing that does feel ordered: sourcing new candidates one at a time, in the order they show up.
The pool already exists — the problem is ranking it
A keyword or boolean search against an ATS answers "who matches these terms." It doesn't answer "who, among the people who match, is actually worth a call first" — a much harder question that depends on things a search field doesn't capture well: how strong the last conversation was, why a past process didn't work out, whether their situation has likely changed, and how close a fit they really are versus how close their resume looks on paper.
Why "just search the database" doesn't produce a ranked list
Search returns matches in whatever order the system defaults to — usually recency or keyword relevance, not fit. Two candidates can both technically "match" a req while one of them is a clearly stronger fit that a recruiter would recognize instantly in conversation and the other is a keyword coincidence. Without a ranking step, both show up looking the same, and the recruiter either reviews all of them manually — slow — or reviews the first few and starts sourcing new candidates to fill the gap, leaving stronger matches further down the list unreviewed.
A practical framework for prioritizing who to contact first
Four questions, applied to each plausible match, narrow a long list to a short one fast:
- Does the core requirement actually match, not just the keywords? A candidate whose resume mentions the right skill in passing is a different prospect than one whose actual recent work centered on it.
- Why didn't a past process with this candidate result in a placement, if there was one? A candidate who withdrew for reasons specific to that role — timing, that client, that team — is a stronger prospect than one who was passed over for the fundamentals this new req also requires.
- Is there a reason to think their situation has changed? Time passing, a title change, or a note from a past conversation about wanting a different kind of role are all signals worth weighing.
- Would a client actually want a call with this person for this specific req, today? This is the question that separates "technically qualified" from "worth using one of a recruiter's limited outreach slots on this week."
Run a list of forty plausible matches through those four questions and what's usually left is a short list of five or six people worth an actual call — not because the other thirty-four were bad candidates, but because ranking surfaces who's the strongest fit for this specific req, right now.
What to do with the ones who don't make the cut
A candidate who doesn't rank near the top for this req isn't necessarily a dead end — the same reasoning that ranks them lower here might rank them higher for a different req next month. Recording why someone didn't make the cut this time (not a fit for this client's stage, right skill but wrong seniority, timing) makes that judgment reusable instead of something that has to be re-derived from scratch the next time their name comes up.
This is triage, not a hiring decision
Prioritization narrows who's worth a call — it doesn't replace the call. The actual read on whether someone's right for a role still comes from the conversation, the client relationship, and a recruiter's judgment about fit that no ranking framework fully captures. The value of triage is protecting that judgment for the candidates who've earned it, instead of spending it evenly across a list that was never sorted in the first place.
Where this gets faster
Running four questions against forty candidates by hand is exactly the kind of task that's simple in principle and slow in practice. That's what Voco does: give it the role you're working on and the candidates you already know — a CSV export of your pool works — and it ranks who deserves a conversation first and shows the evidence behind each judgment, before you spend outreach slots sourcing new candidates instead. Ask Voco.
The same record-trust problem that undermines ATS search generally makes this harder too — a ranking is only as good as the data behind it, which is part of why re-entered candidate data is worth fixing regardless. Once you've ranked a pool, reactivating the names worth another conversation is the next step.