The use of AI is becoming de rigueur in recruiting technology. I love the fact that it’s there, but it’s hard to understand and hard to assess how well it’s working. I was taken with the approach of Kiev-based FourHires which, along with the requisite AI, is stressing the importance of their skills taxonomy for IT jobs (7,000 skills) in order to make precise recommendations when shortlisting candidates.
Taxonomies are an old technology that remain useful. Unlike AI, there is nothing obscure about a taxonomy; it’s easy to see and easy to understand. Easy to understand doesn’t mean easy to build; when I was writing Lead the Work with John Boudreau and Ravin Jesuthasan we spoke with IBM leaders and they discussed how hard they’d been working on building and updating skills taxonomies. Taxonomies are not sexy and not simple, that means their value may be underappreciated.
As you might expect, put together AI and a comprehensive taxonomy and you’ve got a system capable of creating a good short list of candidates. From a recruiter’s point of view, getting that shortlist may seem to be the task that they were meant to accomplish. However, getting a good short list is only half the battle. The other half is convincing the hiring manager that you have the best possible short-list so that they don’t send you back on a futile hunt for someone better.
A good shortlisting technology needs to provide not just a good shortlist, but a shortlist in a form that helps convince the hiring manager that the right talent is on the list. So another tip of the hat to FourHires for emphasizing this part of the process; their report about the shortlisted candidates is aimed at showing managers the quality of the shortlist.
