Why Sourcers are racing the machines

AI is changing the way the world works! Big who ha you might say, but the evidence is in front of us. We are helping the world evolve due to our warm embrace of technology into every part of our lives. Recruitment is the same! Some neigh-sayers fill us with fear that all of our jobs will disappear, others take the time to educate us on how to keep up with the changing landscape. You could believe the apocalyptic science fiction scenarios but my peers and I do not. We design, use and compete against more complex technology every day.

You may ask me why we would want to race against the machines but the reality is it is a way for us to embrace the technology as it evolves. Having competed in the SourceCon Grandmaster vs a Robot, I thought it would be fun to bundle together what losing to a robot taught me about my job and how Robots and AI are helping the world of recruiting evolve.

The problem presented to the SourceCon Grandmaster finalists this year was a simple one, we were given 3 anonymous jobs and a bunch of CV’s and asked; “who got the job?”

The machine finished 3rd and took about 3 seconds, I didn’t.

It’s the problem we face every day, minus all the real work a Sourcer and Recruiter does. I am not surprised that it was so easily and quickly completed by the machine, I was more surprised at its accuracy. It’s an input/output style question: Here is a bunch of CV’s please find me the best ones… yet the reality of what most Recruiters do is more than just throw a bunch of CV’s at a wall and hope that one sticks. Hiring is a subjective choice by both interviewer and interviewee, that’s why robots haven’t taken over.

No one wants to be hired by a Robot Recruiters!

– Hat tip to Matt Buckland who shared this idea with me some years ago.

Controlling the machines was the overarching subject for SourceCon and this will probably be the 100th blog post you read about AI in recruitment. While industry soothsayers continue to amuse us with the threat of replacement, ERE media amused us with presentations of new technology and how to embrace it. I believe that robots will never replace me in what I do. For now, that is because we are not yet comfortable interacting with nonsentient’s in our day to day lives, even though it is now cheap technology, not many people have an automated personal assistant running their lives, yet. As humans, we are still not sure how much we should trust the robots and I’m sure it’s true when we are talking about looking for a job. Personally, I would not be happy if you tried to recruit me with a robot.

That’s why, even with the advance’s technology is making, we are not ready to replace all of the elements it takes to hire people with robots. Sourcing is seen as one of the parts that it can optimise. A miss-understanding that Sourcers are just baby Recruiters is understandable if you have never done either role. From an outsiders perspective, a tool that gives someone who is hiring a list of people makes sense but finding people is easy, especially if you know how to look. A Sourcer and a Recruiters role is more, we all have to engage with potential candidates and have a conversation before we can hire them!

Glen Cathy’s blog was one of those that taught me that finding stuff on the internet is easy, and yet Glen has been working on building and testing these tools coming to replace himself for years. He concluded in his keynote this year that while machines are getting better at looking for people it is the “social engineering’ in recruitment that machines can’t get right. While social engineering is a bit of a scary concept, his slides highlight the humanity that helps us make hires and that what humans do, AI can’t… yet…


Flexing Robot Recruitment Muscles

We like to race against these machines to prove that we are still better. The race and embrace of robotics and AI in recruiting has created smarter Sourcers and Recruiters who flex both empathy and engineering muscles to fill jobs and make hiring people easier. The best Recruiters and Sourcers I know leverage data and automation to fill more jobs quicker. As our industry evolves we evolve with it, it’s not just by wielding the tools that makes us more effective but by also honing the skills we need to use them more effectively that ensures we can continue to bring value to our candidates and our clients.

The role of the Sourcer has moved, we don’t have piles of paper CV’s, we have millions of data touch points readily available. Jeremy Roberts highlighted the shift away from one Recruiter and many Sourcer model to the one Sourcer many Recruiters model and proposed in his talk that there were two archetypes of a Sourcer, Tech or Talker. The truth is that it is the tech that is allowing sourcing teams to be centralised units of experts leveraging data and machines instead of the Recruiters assistant who is ploughing through hundreds of dusty resumes. During the talk, I polled twitter and 69% of respondents felt they leverage technical skills to do their job, while only 8% were left behind with little knowledge on the matter.

I think the response is indicative of how sourcing has evolved, we feel more comfortable using more technical tools to make more hires.

Randy Baily, Sara Goldberg and Guillaume Alexandre BEAT THE ROBOT used in the SourceCon Grandmaster Challenge by using their human ingenuity and gut feelings. While the machine only took a few seconds to match the candidates to the roles. Randy won by leveraging his human ingenuity to take a deeper dive into the data provided. No robot could have done what he did as he went beyond what the challenge asked. The reason why he won is that he took a centaur approach (a phrase coined at the conference to refer to the half man half machine approach to sourcing) No machine can call and convince a human to tell them if they previously interviewed with a particular firm, only a Sourcer with an advanced understanding of all elements of recruitment can. Well Done Randy!

This is what I walked away from SourceCon with.

“Human AND AI” NOT “Human OR AI”

Me plus a machine is better than me on my own. I don’t feel scared by the technology I have seen, only empowered to do my job better.

LinkedIn flirted that 90% of professionals are open to new opportunities. The myth of passive-active candidate (which I never believed in) dies with this statistic as it proves that more people are open to having a conversation. Anecdotally, people are always going to be open to the right conversation if the message is delivered in the right way (Edited – Hat Tip for the comment RB.)

Facebook has 1.23 billion monthly active users, LinkedIn 467 million; with so many potential people to hire.. we need to get better at identifying the right time and the right people to be having those conversations with. We then need to spend time and energy to craft the best approach. Easier said than done. We all have a glut of data available at our fingertips and sifting through it is a painful process. That’s why we need tools to help us do this, that’s what the robots are for, to make us more effective.

I’m not scared of racing robots, they aren’t stealing my job, they here to make me better.

That’s why I raced robots at SourceCon and lost.

If you agree or disagree, please share your thoughts with myself and your friends.

If you want to flex your empathy muscle and learn how to enhance the rest of your sourcing muscles with cool tech. Come work on the bleeding edge of technical and human engineering with the sourcing team Inside Indeed, we are hiring Sourcers!


3 thoughts

  1. Tris, Congrats on being a finalist! Historically, just making it to the final is the really difficult part and something to be proud of! Also, thank you for the shout out.

    There’s one part of the blog that I take issue with and disagree, one of the last paragraphs about the myth of passive/active candidates. While I don’t disagree about the myth part, I do think it is more complex than just saying that the right job will attract the right person. A sourcer needs to be able to get the RIGHT message infront of a candidate at the right moment, ideally when they are in a receptive emotional state for everything to line up. Until we can start to have more predictive insight into those types of things, really we’re just doing another form of post & pray. The thing is there is enough data out there to do this, probably incredibly well, but for the most part it is all locked behind a wall.

    Once we can get to the point of understanding what kind of message an individual will respond to, an idea of when to send it (and I’m not looking for tidbits of data saying that most messaging sent before 6am gets read), then we’ll be able to be that much more efficient. There are some early tools that try to help with some of this, but at this point are hokey at best because they don’t have access to the really good data.

    1. You right of course… maybe I glossed over the subject with some rather broad brushstrokes.

      “A sourcer needs to be able to get the RIGHT message in front of a candidate at the right moment” as you say we have tools to help us with this, but the reality is that for now, it is only a well trained set of empathetic muscles that helps us do this.

      The post and prey method of being able to find everyone and send them all a message is dying a slow death, like my in mail response rate did in 2014. Timing, message, from what and who and where.. there are so many things to consider as a sourcer that a machine can’t yet fathom.

      Looks like you and I are keeping our jobs for another few years!
      (At least until some AI gets there hands on some really good data to train a model on this.)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.