Forbes reports that employers are increasingly using quantitative measures in hiring decisions. (Article here, but free registration required.)
This will scare some folks, but who hasn't had really bad job interviews? One fellow wanted to explain to me his department in detail; he was irritated when I took an interest and asked some questions. He hardly let me talk during the entire interview. In fact, talking too much is a very common error on the employer side of job interviews.
Employers often base hiring decisions on the personality of the candidate, rather than on real qualifications, so moving to quantitative hiring decisions sounds good. But metrics can be used badly, so here are the 4 steps to using metrics in hiring decisions:
1. Collect data on current employees. You need to be evaluating your current employees to figure out what quantitative factors are relevant. Then start going through the data looking for correlations. For example, some companies are using an applicant's credit score in the hiring decision. Don't do that unless you have hard evidence that credit score helps to predict on-the-job success, which could mean performance or retention.
2. Use the data in hiring decisions, either rigidly or flexibly. Flexibly means that the hiring manager can overrule the quants, which is probably best.
3. Evaluate how the hiring rules are working. Oops, don't forget that there are more than two steps. The biggest mistake imaginable is setting up a quantitative system and then assuming that all is well. So here's an approach, using credit scores as an example: you previously found a relationship between credit score and job performance for existing employees. Now you've used credit scores in hiring. Next step is to determine whether your new hires are doing as well as that correlation implies. Are they really doing better than last year's (pre-quant) hires? If not, then go back to square one.
4. Test changes to the system. The system needs to change over time. First, there is no way that you got it perfect the first time round. Second, your company, your tasks, your employees, and the labor pool from which you are hiring are all changing. You don't want to start over completely every year, but you should be testing your rules. What if we relaxed this criterion? What is we tightened that one? What if we only applied this rule if some other measure hits a critical threshold?
Companies that understand quantitative metrics will apply steps 1 and 2 above. Firms that understand the Trial and Error Economy will be applying steps 3 and 4 as well, and they'll be cleaning everyone else's clocks.
Comments