Recruiters and Algorithms: Separating fact from fiction

The subject of “recruiters being replaced by algorithmalgorithms” is a hot topic of debate, and hype, in our industry right now. Check out this article by Matt Jeffery from SAP on ERE. The comment stream alone shows you the level of passion on both sides of the debate.

Those pro the technology suggest that the recruiters job wont disappear, more that it will re orient as technology removes the “mundane elements of their work”, allowing them to focus on other more value adding aspects such as guiding candidates into and through the funnel and improving the quality of the touch points that form the “candidate experience.”

Others question how technology can ever “understand the culture of the organisation” or make appropriate judgments around “fit”. And then, there’s the issue of gathering data. Candidates wont tolerate processes that require them to enter more and more data items simply to apply for your job.

Suffice to say that those who fall under the spotlight will always respond with caution when considering the implications for their own future.

Before we can answer the essential question – can a recruiter be replaced by and algorithm? – it’s worth trying to separate the facts from the fiction and consider the issue objectively.

In order to cut through the hype of algorithms, data science and AI, (In themselves, three very separate but overlapping elements) it’s important to distinguish between the two key implications of this technology in our context. That is the difference between using technology to:

  1. Improve, streamline or remove processes
  2. Replace a human judgement or decision

1. Improving, streamlining and removing processes

Fact: Technology and automation have removed the need for recruiters to screen applications.

Applicant Tracking and Search technology has been streamlining the screening process for years. Typically, these technologies ‘sift’ based on relatively rudimentary structured data points, usually against a framework based on skills, experience etc. In some cases this has removed recruiter heads from the team, reducing overall headcount. In others, it has allowed recruiters to be re allocated to other more value added activity.

More recently, online assessments have grown in popularity, and have had a similar impact. Which brings me onto the second point:

2. Replace a human judgement or decision

Fact: Online assessments have made ‘decisions’ or ‘judgements’ on applicants at the screening level, removing the need for the recruiters own decision making or judgement.

Online assessments are growing in popularity and are shifting the emphasis of screening from skills and experience to values, motivations and behaviours, or a blend of all three. In HR and resourcing parlance, this is often, but not exclusively, referred to as ‘Cultural fit’. These assessments are widely available today and deployed in many large organisations, especially where there are large applicant volumes involved.

Again, recruiters have been removed from the screening decision making process as a result of this technology and either removed from the process altogether or re assigned to more value added activity.

So the fact is, it’s already happening. The question is, what will happen next? The truth is we are still in the very early stages of technology development in this area. Data science and our ability to derive useful interventions from is still maturing. Whilst significant things are happening outside of our industry, the HR/Resourcing landscape generally, is usually one of the last to adopt and benefit from such advances in technology. But it will come, make no mistake about that.

Top put it into context, I will refer to a comment made by industry veteran John Sumser back in December 2011 when discussing the development of social technology and its impact:

“We are in the Compuserve-Prodigy-AOL stage of social media evolution.  It’s after Netscape and before Google in equivalent internet time.”

In data science terms, we are still very much in the equivalent Compuserve/AOL time, possibly even earlier, hence our inability to see exactly how this will impact our industry. Whilst technology is getting smarter, the majority of solutions are still using skills, experience, location etc. to identify individuals and position them as good for a role, or at the very least, “rank” them in order of suitability.

The emerging technology may look at this too, but the game changing stuff is also looking at a different set of data – your digital “footprint” data – in the form of what you do, how you do it, what you have said, your comments, Facebook likes, status updates, LinkedIn updates and so on.

In fact, all the data that is needed to predict, with a significant degree of accuracy, your potential and your propensity to behave in a certain way, is pretty much already out there. And it’s already being traded and used in order to make decisions about you. Whether that be your qualification for a loan, your likelihood to buy something or your insurance risk. It is only a matter of time before the capability finds its way into our industry.

Essentially it means taking what assessments currently do – assessing you via questionnaires – and instead assessing you “frictionlessly” (You heard it hear first folks!). That means assessing your values, motivations and behaviours without you having to enter any data, or complete a questionnaire.

The sceptics among us throw up many arguments to justify why “it will never happen” including some I found below from recent articles:

“How can an algorithm gauge the non-skill related qualities a candidate needs to succeed in the role such as passion, dedication, likeability, EQ, potential, and learning agility?”

“Is there a program out there that will capture what the corporate culture is? How will these tools measure that?

“No matter how automated it is, someone (candidate) still needs to input a load of data / information, and the more the algorithm wants to assess, the more information will need to be inputted.”

All of these points may seem valid now, but they miss the fundamental shift that is happening in technology. Take the last one for example – frictionless assessment doesn’t require any data input at all. The only thing it requires from the candidate is permission.

Think about it. Elsewhere fundamental and commercially very important considerations and decisions are being supplemented and in some cases taken, by technology. Are we really saying that this sophistication will never reach our industry?

So for me, the question is NOT will this approach replace the recruiter. It’s how much MORE of the process, and judgement of the hiring person can be augmented and or replaced by the technology.

Methinks, quite a lot.


  1. I might turn this on its head.

    The implication in here is that for each recruitment event (candidate is matched with job) some data is gathered anew, a profile created and matched to a desired profile (which may be right or may not, how often does one interview some one and find, not what you were looking for but something you then know you want).

    If we observe that recruitment is a form of sales process but a two way one.

    The recruiting org has a culture profile and a need profile. At least it thinks it does.

    The candidates each have a set of preferences, behaviours, skills, circumstances and needs at a point in time. At least they think they do.

    Stepping outside recruitment and into more normal market making, a seller buys a volume of analogous personal profiles from FB, LinkedIn, Yahoo or whoever, match preference on both sides and engage in various flavours of match making. Nothing revelatory in that except that each campaign does not involve in any new data gathering, the currency of the profile flows without friction (unless you count the annoyance it causes).

    This potentially changes process and begs bigger questions.

    Can profiles be standardised and enriched to be useful on both sides?

    Who owns them ? One side is a rich record of identity. 

    Who controls them (am I looking, who can see that)?

    What other valuable data points exist? Connections, references, reputation, opinion. If we can match the basics quickly, do these types of data become more valuable, much much more valuable. 

    Do King makers emerge? Well connected people whose reputation is good and whose opinions are as valuable as the cost of a bad hire?

    When might I ask a candidate for more information ?

    Could I assess the fit of a million or a billion candidates for any role?

    Does role often quickly morph into task?

    As so often with digital, the opening statement becomes, “I wouldn’t start from here”.

    The process doesn’t change, it is completely different.

    Liked by 1 person

  2. Hi Anthony! Long time no speak. Maybe this is one for a beer!

    You speculation around buying profiles and doing matching is entirely possible, but there are a couple of roadblocks that make this hard to do at the moment. 1. You need a proper view of what good looks like for a specific role in a specific organisation. Talent is situational, so without that, its a blind match. 2. Creating insight from such profiles requires a lot of heavy lifting in terms of building the models in order to do the matching or predicting. 3. This data gathering and modelling is expensive, very expensive. 4. Permissions. Its going to get more difficult to get access to profile data. 5. Supply – some of the more valuable “unstructured content” will become harder to obtain/be more scarce.

    On the point of king makers, reputation will definitely become more important – there are already businesses in this area that are pivoting to reputation models – but in a big data sense, the bigger a reputation you have really has no impact. There are many reasons for this, probably too many to list here.

    I can feel a pint coming on…. 😉

    Liked by 1 person

    1. Hi

      Beer, definitely 🙂

      In the middle of all that was a question of ownership.

      If my profile is mine (not owned by some platform) and stored on a BlockChain, then the game changes.


      See also the value of the King maker, the value flows to them.

      I talk of ownership and currency quite deliberately. The current platforms are toast.

      Different sphere, similar argument, but so are über and Airbnb unless they move into new models quickly.

      My round ?


      Liked by 1 person

      1. Sorry, skipped the ownership question! Agree totally, its a game changer. I believe we are heading into the ownership issue anyway, outside of what we have discussed here, even at the typical enterprise ’employee record’ level. So even without the prospect of blockchain, i see real challenges for the existing enterprise platforms because they are not set up to manage that ownership challenge. I’ve eluded to this in a previous blog –

        Put blockchain into the mix and potentially the whole world could shift on its axis. This could, potentially, be good news for the individual. The control and value shift to the individual could be very powerful. It also raises some interesting questions/issues too. For example, with the principle of ‘validation’ within blockchain, how do i feel with effectively losing “control” over how my validation, and value, is defined? At the moment, i kinda still self validate, which means i can, to some extent, be who i want to be. Present or mould the face i want to present in order to achieve my personal goals. Career goals in this case, or as a ‘trusted’ consumer in others.

        Fascinating stuff. Reckon we have enough material there for several rounds 😉 I’ll ping you some dates for a natter.

        Liked by 1 person

  3. I remember your HR record post. I’m surprised I didn’t comment 🙂

    I recall vividly where I was when BlockChain hit me. I was on the downs outside Winchester and listened to an economist podcast. I had to stop and sit down. It may not actually turn out to be that but something like it.

    I eventually scribbled this :

    Any way, hols next week. Otherwise, take your pick. I’m in London and it’d be good to see you.


  4. 35 years I spent in HR.
    If I had a dollar for every time I read another Industry GURU going on about replacing Recruiters..
    Technology has it’s place in the grand scheme and should be leveraged however “Artificial” intelligence will never replace the human brain. At the end of the day Recruiting is about people and judgments made within the scope of human interaction.



    1. Hi RWS!

      Thanks for reading the blog and your comment. As someone who has spent almost as much time in the HR and recruitment industry – 30 years as a practitioner and consultant – I too have seen a lot of hype around this. The point is though that technology definitely has a role to play and it will increase. In some parts of the resourcing process, it is far more effective, reliable and objective than human decision making. Thats just a fact. Humans are biased, and so are recruiters, very much so. We cant help it.

      Unfortunately when this subject is debated, many seem to leap straight to the conclusion that we are saying the recruiter will be completely replaced and that a new hire will simply turn up on day one without ever meeting a human being. That is not what i or anyone else is saying but we seem to react badly and go straight to extremes. In my post above, i talk about assessments replacing recruiter time in screening. This is happening and should happen. I don’t know why recruiters would not want this – i would have done in my days of recruitment, especially in high volume markets. Unless of course they don’t value their role further down the resourcing pipe and consequently ‘hide’ in being busy scanning CV’s!

      I agree and think its important not to jump on the hype. But i also think its about time we bought efficiency and objectivity to the screening process which technology and assessments do.

      Thanks for stopping by.




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s