In classification models, there is a concept of precision (% of correct positives) and recall (how many positives are captured from the population). The author

In classification models, there is a concept of precision (% of correct positives) and recall (how many positives are captured from the population). The author correctly points out that recall is very low in a typical recruitment process. As much as I agree with the sentiment, employer incentives are not to find ALL suitable candidates but to find just ONE suitable candidate. Everything beyond that is a waste of resources. For some positions (e.g. finding a co-founder or an artist with a unique style) - yes, low recall might be a limiting factor. It is just a market for most other roles - in particular, all that can be summarized with a job title.