Automate sourcing: what works and what doesn't

Automating sourcing is handing off the groundwork: the searching and getting a first selection ready. A system reads your vacancy, turns it into search criteria and pulls the profiles that fit. It drafts a first version of your message, written for that one person. You see the list, drop what does not belong and adjust the search when it misses. Within a few minutes you can tell whether the criteria are right: too broad and you get a hundred names, too narrow and the list stays empty. Tuning that is your job, and it takes seconds.

What it is not: messaging candidates on autopilot, or a tool that promises you never have to search again. Those tools exist. They send the generic messages that good candidates spot and ignore. The difference is in who does what. The AI does the searching and the prep; you decide who you actually approach and what you end up writing.

Before you open a new channel, look in your own ATS first. That is where the candidates sit who you spoke to last year and who did not quite fit at the time. For today's role they might fit fine, only nobody looks there, because manually digging through thousands of old profiles is not realistic. A system does it for you. It reads the new vacancy and pushes the people from your own records who come closest to the top, whether they live in Bullhorn, Carerix or OTYS.

Old data is often half filled in or out of date, and a good system accounts for that: it shows which details you still need to check before you call. Take a request for a financial controller in Utrecht. Instead of starting on LinkedIn again, the system runs the question past your own records and returns eight people who already came by two years ago. Three are still worth a call, and nobody else would have found them that day. This is where we at Plaatsr put your existing database back to work for sourcing, as one of the systems we build. No new source and no wait, just the people you have already spoken to.

If your own records come up short, then go outside, to LinkedIn and the places where your audience actually hangs out. There too, automation helps with the heavy part: the searching and keeping track of who you found where. Without an overview, sourcing across several channels quickly falls apart. You approach the same candidate twice, or you forget to follow up with the one person who was interested.

For Match Masters we built a dashboard that makes the step before the ATS clear: candidates and follow-ups in one place, where they used to sit in scattered lists. Steijn Tersteeg sees on one screen who he found where and who he still needs to follow up, and gets roughly two hours a day back that used to go into tracking. Read the Match Masters case. New channels only cost you time once you lose the overview. Let the system guard that, and you keep your hands free to reach out.

Automation pushes you toward volume. One click to message two hundred people is tempting. It still does not work. Candidates spot a mass message instantly, and the very people you want do not reply to it. Volume feels like progress because the numbers go up, but your response rate drops and your name wears thin. The gain from automating sourcing is not more messages, but better ones. The system hands you a short list and a first version that fits that person; you finish it with the details only you can see. Keep an eye on which kind of message gets a reply and which does not, and the system learns to sort more sharply next time. A few sharp messages do more than two hundred generic ones. And a sloppy message sticks: that is how a candidate ends up remembering your agency, for the wrong reason.

The conversation. The judgement call on whether someone with a not-quite CV is still the right person. A system sees patterns in text, not people. It does not know that a candidate has just become a parent and does not want to travel, or that someone who looks overqualified on paper is ready for a step back. You hear that in a conversation, not in a profile. Automating sourcing mainly gives you the room to have that conversation more often, because the searching no longer swallows your whole morning. The groundwork is for the AI. The judgement and the contact stay yours.

Sourcing is one step. The CV screening after it and the submittal: there is just as much time to win there. To see how those steps connect, read the guide on automating recruitment. Sourcing is usually a good place to start, because you feel the result straight away: a shorter list and a first contact that stands sooner, without the hours of searching up front.