How to source passive candidates with AI.

Why job boards can't find your best candidate

The best superintendent in your market is running a $60M job right now. He hasn't updated his resume since 2019, he's never had an Indeed profile, and he ignores recruiter InMails on principle. Every placement worth a real fee looks like this. The people your clients actually want are working, not applying.

That's the structural problem with sourcing tools built on applicant flow: they show you the fraction of the market that's actively looking, which is precisely the fraction your client didn't need you for. Passive sourcing means searching the whole market, and the whole market only exists in profile data.

What plain-English AI search actually changes

Boolean search made recruiters half-engineers. The craft became writing the string, not knowing the market. Plain-English sourcing flips it back: you describe the person the way a client describes them on a kickoff call.

In RecruiterClaw that looks like typing into Slack: "Find me superintendents in Dallas with ground-up commercial experience, 10 plus years, currently at a GC." The AI turns that into a structured search across 400M+ profiles, shows you its plan before spending anything, and returns a ranked shortlist in about 90 seconds. The ranking isn't a black box either. Ask why someone scored high and you get the evidence quoted back from the profile.

Refinement is a conversation, not a rewrite: "drop the estimators," "tighten to 30 miles," "only companies like the ones on my target list." Each pass sharpens the pool without starting over.

Only pay for contact data on people you actually want

Here's where most sourcing stacks quietly bleed money: they enrich everyone, and you pay for verified contact data on 200 people when you were only ever going to call 15. The discipline that matters is scores first, spend second.

RecruiterClaw holds contact enrichment until you ask. Review the ranked list, then say "enrich the top 5" (or 15, or 50). Verified emails plus direct and mobile phones come back appended to your CSV for exactly those people. Nothing burned on candidates you never vetted.

Finding them is half the job. Opening them is the other half.

A passive candidate who picks up the phone and hears a job description hangs up. He's happy, or at least he's decided he's happy, and a pitch confirms he should stay put. The opener that works is a question, not an offer.

This is why RecruiterClaw ships with the Field Manual: a recruiting playbook built from 30+ years of desk experience, covering the passive-candidate discovery arc, objection handling on both sides of the desk, and counteroffer inoculation. Ten minutes before the call, ask your Chief of Staff how to open a superintendent who says he's happy at his GC, and it coaches you from the manual, adapted to your niche. If you've built your own playbook over the years, upload it; yours takes precedence.

What this looks like as a daily system

Sourcing passive candidates isn't one search, it's a rhythm. A search when a job order lands. A refinement pass after the client call changes the spec. An enrichment run before Thursday call block. RecruiterClaw's flat pricing matters here: tasks, conversations, and scheduled automations are unlimited, so you never hesitate to run one more search on a hunch. Only the sourcing data itself draws from a monthly allowance, and it's sized generously.

One founding client, Randee, made two placements worth more than $40,000 in her first month, from candidates she says she never would have found. That's the compounding math of searching the whole market instead of the active sliver.