HR Has Enough Best Practices

Last night, after spending a couple of hours onstage discussing recruitment, AI and the future of HR, I found myself in a conversation with a group of people who had stayed behind for the networking.

One of them, a Head of HR, was challenging me on something I had said during the panel. They thought I was contradicting myself. We had been discussing whether AI could assess candidates, and they were adamant that it couldn't understand the passion or conviction with which someone speaks about their experience.

Someone else in the group started to intervene, telling them to take it easy on me.

I asked them not to.

This was exactly the kind of conversation I had hoped we would have.

Before the event, I had encouraged every panellist to challenge me whenever they disagreed. I had given the moderator the same invitation. I'd even asked people during the networking beforehand to push back on what I said. We had promoted the evening as a no-bullshit HR conversation, and I wanted us to mean it.

There are already enough professional events where everyone politely agrees with one another. We listen to familiar principles, congratulate ourselves on sharing the same values and go home feeling reassured. It can be pleasant, certainly, but I'm not convinced we learn very much.

So when this HR leader questioned me, I wanted to understand their argument.

I asked them what they meant when they said AI couldn't recognize passion and conviction. What, specifically, would a human interviewer pick up that an AI system would miss?

They explained that they wanted candidates to describe real situations, what they had been asked to do, how they approached a problem and what results they achieved.

As they spoke, I realized they were describing something remarkably close to the STAR method, a familiar way of structuring answers around a situation, task, action and result.

And that gave us somewhere useful to go.

I explained how we use structured interview scorecards. Before an interview, we define the competencies we need to evaluate and what different levels of evidence look like. What would make an answer a four out of four? What would make it a one? What would we expect a candidate to describe, demonstrate or explain?

Once you've done that work, those criteria can also be given to an AI system.

I could see that the idea of applying scorecards to AI assessment was new to them. We had moved from a fairly broad disagreement about whether a machine could understand a human being to a much more specific discussion about what we were actually trying to evaluate.

And that's where the conversation became interesting.

I don't know whether AI can reliably assess someone's passion from their tone of voice or the conviction with which they tell a story. I'm certainly not prepared to claim that it can. But I'm not convinced that passion, or at least the ability to perform passion in an interview, should be a hiring criterion in the first place.

Someone who speaks confidently, makes excellent eye contact and tells a compelling story can leave a wonderful impression. That doesn't necessarily mean they can do the job better than someone who's nervous, less articulate or simply uncomfortable talking about themselves.

If we can't explain the difference between what impressed us and what demonstrated competence, perhaps the problem isn't that AI lacks human judgment. Perhaps it's that we haven't been particularly clear about what our human judgment is based on.

I run into a version of it regularly in my work.

A hiring manager finishes an interview and tells me they just didn't feel it with the candidate. I understand that reaction. We all form impressions of people, often very quickly, and those impressions can be difficult to put into words.

But my response is usually to ask why.

Was it the way the candidate dressed? The way they spoke? Did they avoid looking at the camera? Were their answers too high-level, or did they fail to provide relevant examples? Was something missing that mattered for the role?

Those are very different explanations, and they don't all deserve the same weight.

My job isn't to tell a hiring manager they're wrong to have an instinct. It's to help them examine that instinct before it becomes a rejection.

We have worked on recruitment assignments where every candidate decision needed to be documented and justified because the hiring was supported by public funding. There was no room for a vague explanation that someone simply didn't feel right. We had to be able to articulate why a candidate had progressed or been rejected.

I've always thought that was a discipline worth applying to every recruitment, whether someone requires it of us or not.

That doesn't mean our understanding of the ideal candidate has to remain frozen from the moment we write the job description. Sometimes we discover halfway through a search that a particular skill matters less than we thought, or that another quality is far more important. Recruitment is a process of learning, and it would be foolish to pretend otherwise.

But if our expectations change, we should be able to explain what we've learned and make sure the people involved in the decision are aligned.

It's the same principle I apply when trying to understand a client's working environment.

During the panel, we discussed transparency and how honest we should be with candidates about difficult managers, demanding cultures or the less attractive realities of a job.

For me, honesty starts well before we present someone with an opportunity. It starts with being curious enough to find out what the organisation is actually like.

As external recruiters, we see a company from the outside unless we're invited, or ask, to look further in. A polished briefing and a convincing description of company values are useful, but they're hardly the whole story.

I pay attention to how clients work with us. Do they show up for our regular conversations? Are their expectations consistent? Can they explain their decisions? When something goes wrong, do they work with us to understand it, or immediately look for someone to blame?

I've spent time inside one particularly demanding organisation we recruit for. I watched a co-founder sit down with the engineering team over lunch and work through a difficult problem, asking people for their ideas rather than simply telling them what to do. I also spent time with employees without the founders present and asked directly how they felt about working there.

The feedback was overwhelmingly positive. Many had made significant changes to their lives to join the organisation, and they knew what they were signing up for. They believed in the work, enjoyed the challenges and felt that their contributions mattered.

It was an intense environment. I'm not going to pretend otherwise. But I'd seen enough to feel comfortable explaining that intensity honestly to candidates, rather than treating it as either a selling point or a warning label.

I've also worked with clients who made the recruitment partnership unnecessarily difficult, treated our expertise as a disposable support service and seemed determined to cut corners or shift blame whenever things didn't go as planned.

We generally don't work with those organisations for very long.

There is a difference between demanding work and a workplace that doesn't respect the people doing it. I'm not sure you can discover that difference through a company presentation. You have to ask questions, observe behaviour and sometimes challenge the people paying your invoices.

One discussion during Thursday's panel troubled me for a different reason.

We were talking about what candidates want today. Flexibility, remote working, work-life balance and the freedom to choose were all part of the conversation.

Those things matter. I've built much of my career around the belief that candidates deserve to be treated as people with lives, needs and ambitions of their own.

But as I listened, I felt we were missing an enormous part of the picture.

We were largely discussing candidates who have choices.

There are people who can afford to compare employers, negotiate flexibility or turn down an opportunity because the working arrangement doesn't suit them. And there are people who have been unemployed for months, sometimes much longer, for whom the most pressing question isn't how many days they can work from home. It's whether someone will give them a chance to work at all.

I've met people at recruitment events who have gone a year without finding employment. Some have sought me out again at later events, asking, almost pleading, for help.

It's difficult to forget those encounters.

I think my sensitivity to this divide goes back much further than my career.

I had an extraordinarily privileged upbringing in Canada, and I've always known it. Both my parents had successful careers. My father's work in international basketball gave us opportunities to travel that most families simply wouldn't have. My mother, a teacher and educator of teachers, was also deeply involved in welcoming refugees who arrived in Canada.

As a child, I sometimes found myself looking out from the comfort of a hotel at children working on the streets. At home, I was exposed through my mother's work to stories of families who had fled circumstances I could barely imagine.

I was witnessing very different realities, but always from the protected position of someone who could return to a comfortable life.

I struggled to find work for a few months after completing my master's degree. I'd invested years in my education and hoped for something that would allow me to build a career, but my immediate options were limited. Recruitment eventually presented itself as the first real opportunity, and what began almost by circumstance became the profession I've spent my working life in.

That experience was frustrating. It was not the same as the circumstances of someone who has no financial safety net and is desperately trying to find employment.

When we speak about what candidates want, I think we owe it to them to remember that they are not one homogeneous group. Their circumstances shape the choices they can make, and some have far fewer choices than others.

Acknowledging that isn't an argument for accepting poor working conditions. Quite the opposite. It should make us more careful about the promises we make, the opportunities we present and the way we treat people whose need for work might make them reluctant to ask difficult questions.

There's an interesting contradiction in all of this, one that became particularly obvious as my week unfolded.

On Monday and Tuesday, I attended training on building AI agents. During one exercise, I created an interview agent in about an hour. It could listen to what I said, transcribe my answers and evaluate their relevance against criteria for a role I was recruiting for.

It worked surprisingly well.

That doesn't mean I had built a production-ready recruitment system, or that I'd proved it could evaluate every candidate fairly. Those are much bigger questions, involving testing, transparency, candidate experience and careful human oversight.

But I had built something functional in the space of an hour that would have seemed fairly extraordinary not so long ago.

Two days later, I was sitting on a stage discussing whether AI could play any meaningful role in assessing candidates.

I couldn't help thinking that if more of the people in that room had been beside me during the training, we might have been having a different conversation.

Instead of spending so much time debating whether AI could do something in principle, perhaps we would have been asking what we could build, how we would test it and which problems were worth solving.

One panellist repeated a line about America innovating, China copying and Europe regulating. It's a memorable line, still, the frustration behind it resonated with me. Here in Luxembourg, I sometimes feel we're having conversations about technology as though its possible applications remain hypothetical, while other people are already experimenting with them.

I mentioned a candidate who had been comfortable progressing through an entire recruitment process without speaking to a human. What mattered most to them was the opportunity itself: flexible working, fair compensation and interesting work.

I wouldn't take one person's experience and turn it into a universal rule. Some candidates would hate such a process, and we'd be foolish not to listen to them. But the example challenges a common assumption that every candidate necessarily wants more human interaction at every stage.

Maybe they want an opportunity to demonstrate what they can do. Maybe they want clear information, fair treatment and a timely decision. Human contact can help deliver those things, but its presence doesn't guarantee them.

That brings me back to the HR leader who challenged me after the panel.

I don't think the point is to replace human judgment with an algorithm and declare the problem solved. Nor do I think the answer is to defend every existing human-led process simply because it involves people.

We need to understand what we're assessing, why it matters, how we know whether we're right and what happens when we're wrong. Only then can we have a serious conversation about which parts of a process should be handled by people, which might be supported by technology and which perhaps shouldn't exist at all.

The conversation continued on today, at a small lunch gathering with other HR leaders. Several had attended the panel the night before, so the discussions picked up almost where we'd left off.

A few people described me as representing a new generation of HR, in contrast to their more traditional generation.

It was a strange thing to hear.

These are people I've spent nearly ten years learning from. When I arrived in Luxembourg, I went to events like these because I wanted to understand how experienced HR leaders operated, how they influenced their organisations and how they made the transition from a support function to a genuine business partner.

I still have plenty to learn from them.

And yet we were now discussing whether some of the practices we've all grown accustomed to might need to be reconsidered entirely.

During the lunch, we were asked to introduce ourselves and share a challenge we were working on. I explained that I was trying to think beyond simply adding AI to our existing recruitment processes.

What would happen if we were willing to take those processes apart and rebuild them?

Not because everything we've done until now is wrong, or because every technological possibility deserves to be adopted. But because new capabilities give us an opportunity to ask questions we might not otherwise have considered.

Why do we use a CV as the primary entry point when it tells us so little about what someone can actually do? Why do we rely on unstructured impressions when we could define and evaluate relevant evidence? Why do we make candidates repeat themselves through multiple interviews? And why are we so attached to particular steps of a process when we sometimes struggle to explain the value they add?

Of course, none of this happens in isolation from the business.

Someone recently told me that the HR function at their organisation had lost its seat at the executive table. I found that profoundly disappointing, particularly at a time when decisions about technology, skills and the future of work are going to reshape organisations.

But I also think HR has work to do if it wants to influence those decisions.

We need to become better at explaining our value in terms that make sense beyond our own profession. Words like humanity, culture and candidate experience matter, but we can't expect a CEO or CFO to understand their business impact simply because we believe in them.

We have to make the value tangible.

That might mean demonstrating how better assessment improves the quality of hiring decisions, how clearer processes prevent costly mistakes, or how a more thoughtful candidate experience helps us reach people we'd otherwise overlook. It also means being prepared to measure outcomes, admit when something isn't working and change course.

If we're asking to be taken seriously as strategic partners, curiosity and good intentions won't be enough. We have to show what changes because we're there.

Which, strangely enough, is not so different from what I ask of a candidate in an interview.

Tell me what you did. Explain why you did it. Show me what happened as a result.

I didn't leave last night’s event having changed everyone's mind about AI, and that was never my intention.

I'm not sure I changed my own position on anything fundamental, either. What I gained were more perspectives, better questions and another reminder of how differently people can understand the same profession.

I also left thinking about the HR leader who had challenged me and the person who'd tried to protect me from that challenge.

It would have been easy to end the conversation politely, acknowledge that we saw things differently and move on to someone else. Instead, we stayed with the disagreement long enough to understand what was underneath it.

That, to me, was the most valuable part of the evening.

HR has plenty of frameworks, best practices and reassuring language about putting people first. We can all repeat them. We can even genuinely believe in them while continuing to make decisions based on habits we rarely examine.

What I would like to see more of are conversations where we're willing to explain those habits, challenge them and let other people challenge ours.

Not to win an argument, and certainly not to pretend that every difficult question has a neat answer.

But because the work matters too much for us to keep agreeing with one another out of politeness.

And because sometimes the most human thing we can do is ask someone, with genuine curiosity, why they believe what they believe.

If you’ve made it this far, thank you. Sincerely. I’m grateful you’re here.

If anything made you think, react, or question, let me know!

Until next time 🍁