Ask most agency recruiters how they match candidates to new job openings, and you'll get an honest answer: memory. Not their ATS. Not their CRM. Their brain. That works fine until the day it doesn't, and by then the candidate who would have been a perfect fit is already placed somewhere else.
Why do agency recruiters keep candidate matches in their head instead of their system?
Because most systems were built to log activity, not to think. Your ATS tracks stages. Your CRM tracks contact history. Neither one tells you, unprompted, "this candidate you screened eight weeks ago is a perfect fit for the req that just came in." So recruiters do that work themselves, in their heads, because nothing else is doing it for them.
Adam Posner runs NHP Talent Group in New York. He's been recruiting for 11 years. When I asked him how he tracks which candidates fit which roles, he said it was a spreadsheet. Then he corrected himself. It was actually two spreadsheets. One in his system. One in his brain.
That's not a knock on Adam. That's the reality for most solo and small agency recruiters. You are the database. You remember who you talked to, what they wanted, which client might want them. The spreadsheet exists. The real matching happens between your ears.
What happens when a recruiter's memory is the only database?
It works right up until you get busy, and then it quietly starts costing you placements. A role comes in on a Tuesday. Somewhere in your database is the exact candidate for it, someone you screened six weeks ago and never followed up with. You don't remember them. Nobody prompts you to look. A recruiter at another agency finds that same candidate first, and the placement goes to someone else.
That's not a discipline problem. Nobody can hold hundreds of candidate profiles, skill sets, and conversation histories in working memory and reliably retrieve the right one on demand under deadline pressure. The failure isn't the recruiter. It's expecting a brain to do a database's job.
How does AI candidate matching software close the gap?
It runs the search you don't have time to run yourself. The moment a new req comes in, Agency+ scans your entire CRM and surfaces the candidates who actually fit, not just by job title, but by skills, past roles, and what you noted from prior conversations. You're not relying on remembering a name from two months ago. The system already found them.
This is what the Conduit layer underneath Agency+ is built to do: turn a static candidate list into something that actively thinks alongside you. Pair that with the ATS+ workflow you're already running placements through, and the matching happens as a byproduct of using the system, not a separate task you have to remember to do.
Adam's a one-man shop. By his own admission, his whole database used to live in his head. His take on tools like this: "If you're not evolving as a recruiter, you're going extinct."
What's the real difference between an ATS and true candidate matching?
An ATS tells you where a candidate sits in a pipeline. Candidate matching tells you which candidates, out of everyone you've ever talked to, fit a job that just opened. Those are different questions. Most recruiting software only answers the first one. Agency+ was built to answer both, because a placement depends on finding the right person just as much as tracking them once you have.
Frequently Asked Questions
Do I need a large candidate database for AI candidate matching to work?
No. Even a database of a few hundred candidates benefits, because the problem isn't volume, it's recall. A solo recruiter with 300 candidates in their CRM still can't reliably remember who fits a brand-new req on demand. Matching software solves that at any database size.
Does AI candidate matching replace a recruiter's judgment?
No, it removes the recall problem so judgment can actually get used. The software surfaces who fits based on skills, history, and role data. The recruiter still decides who to call first, how to position them, and whether they're right for the client. Agency+ was built to hand recruiters better options, not make the decision for them.
How is this different from a basic keyword search in my ATS?
Keyword search only finds what you type. If you search "account executive" you'll miss the candidate whose resume says "AE" or lists relevant skills without the exact title. Real matching looks at the substance of a candidate's background, not just whether a string of text happens to match.
Can a solo or small agency recruiter use this, or is it built for big teams?
It's built for agencies of any size, including one-person shops. Adam Posner runs NHP Talent Group solo, and the memory problem he described is actually worse for small agencies, because there's no second recruiter to catch what the first one forgot. PerfectHire built Agency+ around that reality.
How fast can I see which candidates fit a new job opening?
As soon as the req is entered. Agency+ scans your existing CRM immediately and surfaces fits without you running a manual search. You can see recommended candidates before you'd have finished trying to remember them yourself.
If your best candidates are living in your memory instead of your system, you're one busy week away from losing a placement you already had. See how Agency+ finds them for you.