Choosing who to hire is one of the world’s oldest problems, and even the latest technology doesn’t help.
AI in recruitment? No thanks.
While we may be getting used to AI taking a role in the recruitment process, many of us don’t like it. Research from Greenhouse, a hiring platform, found that 30% of UK candidates had withdrawn from a job opportunity because an AI interview was part of the process. And of those who did continue with an AI interview, many reported on the experience negatively.
Sorry, you failed the secret test…
But the alternative approach can go too far the other way… Duolingo was recently in focus for using feedback from the candidate’s taxi driver to decide not to appoint a candidate for a very senior role. There are many similar examples from other companies – such as whether someone offered to take a mug back to the kitchen, trying to identify how people really interact, and the values they actually display, outside of the interview room.
It’s a difficult choice.
In both cases, the companies are trying to make a difficult choice, one that can have lasting and significant impacts. It’s not surprising they are looking for ways of making a complex process easier and more effective for them, whether it is using AI for initial screening that might take a person days to accomplish, or at the other end of the spectrum, trying to answer that intangible question about who is right for the company.
It’s a 10 from me too!
And companies keep looking for new approaches. While recruitment has always been a competitive process, some firms are fully leaning into this idea with recruitment campaigns that resemble TV talent shows. Companies seek to reach a new labour pool who might not have considered applying via the traditional route – L’Oreal receiving 380,000 applications from global candidates who had to develop and pitch ideas for the chance of a job.
So for me, ‘who to hire?’ is not the biggest question in recruitment – instead, you should focus on ‘how to hire?’
If you get the process wrong, your AI might filter out some high-quality candidates who don’t respond well to the AI interview, but could have performed the job well.
Or you could end up making a vital decision based on an assumption or misinterpretation of someone’s behaviour. You could find yourself open to claims of discrimination and bias for a subjective process.
Or you build a recruitment process that requires candidates to invest a huge amount of effort with little chance of reward – thereby excluding people without time or resources. Or the process favours a certain type of person – which isn’t necessarily linked to their suitability for the role.
I regularly advise on recruitment processes and train managers on the skills they need to make good recruitment decisions, and there are three themes I focus on:
Transparency, fairness and consistency are the keys to good recruitment
Whether you are using AI as the only feasible way to fairly screen hundreds of applications, or using tests to assess real-life interpersonal communication, you need to make sure that:
- your candidates know what is expected of them, and they know how they are being assessed
- the assessments are clearly linked to the job, and the process inadvertently make it harder for certain groups of people to succeed
- the processing, testing, and assessing are structured and repeated so that all candidates are judged against the same criteria and the same circumstances.
After all, isn’t that how you would want to be assessed?
