How do you do it? Is there a right way to do it? Does it all boil down to the famous saying attributed to Steve Jobs: “A people hire A people, B people hire C people”? Obviously there is no easy answer to this question, but we can share some bad news and some good news in that respect.
First the bad news. Yes, you guessed it: there is no way guarantee or even near guarantee a good hire. In our previous discussion about predictive validity of different pre-employment evaluation techniques we showed that every single hiring method, however solid it might look, is inherently inaccurate. Some, of course, are better than others. If you hire somebody with whom you worked before and know from firsthand experience to be an excellent performer, the odds are very good (yet still not perfect). From this ideal scenario the odds quickly go downhill. You know it when your learn that the added accuracy of a trusted hiring tool such as structured interview can be a mere 10%…
Now the good news: If you are disciplined and organized in your hiring processes, you can maximize the odds of making the right decision. All you need to do is to use the most powerful evaluation techniques as early as possible in the process. When a candidate comes top ranked on multiple evaluations, it is less likely that any one high evaluation was a fluke. Of course it’s important to understand that combining an evaluation with predictive validity A with an evaluation with predictive validity B does not necessarily give you predictive validity A+B, however it can often give a predictive validity higher than A or B alone.
Combining multiple evaluations might at first look like a practice with diminishing returns. Indeed if you were to apply all evaluations to all candidates it would be costly and time consuming. And here is what makes this good news really good: you don’t need to do it. You don’t need to apply all evaluation techniques to all candidates. As long as you start with the evaluation with the highest predictive validity and go to the next one with the second highest and so on, you can focus in every stage on only the top of the list group.
Cangrade personality assessments have the highest predictive validity of all assessments known to us (for full disclosure: I’m the CTO of Cangrade). How do they work? We have a full section on our site dedicated to it, but to give you a very basic idea: we collect data from hundreds of thousands employees in the US about their performance along with different aspects of their personality. Then we build a mathematical model that connects these two variables. Obviously we take into account many different personality attributes and a multitude of job profiles. The personality data is collected by administering psychometric assessments, and performance data is collected from multi-source performance reviews. When a candidate takes Cangrade personality assessment he or she just needs to complete the same 15-minute assessment that previous candidates and employees also completed when we created to predictive model.
The second highest predictive validity known to us is skill testing. At Cangrade we provide a variety of skill tests, and you can also add your own. These tests can be conducted online and are very cost-effective and easy to administer. Some companies choose to conduct skill testing in person (e.g. white boarding and in person Q&A) either to reduce potential cheating, or to understand how the candidate works under pressure or around others. With either approach, it is essential to consider the utility of the skills being tested–is this something that is needed in a candidate, or something that can be picked up with onboarding, training, or experience on the job? An in-person structured interview comes next in the order of validity. We will dedicate a full post next week about DOs and DON’Ts of an interview. Stay tuned.
When you use the most effective assessments it may be possible to achieve a validity of 50% or more in your ability to predict a good hire. This may not sound high to you, but keep in mind that some amount of inaccuracy is to be expected, and may not be particularly consequential to your hiring process. For example, if an assessment predicts that a candidate will perform in the top 1% and they turn out to perform in the top 5%, the prediction wasn’t perfect; however, most of us would still consider the candidate to be a top performer, and certainly a “good” hire rather than a “bad” one. If you are not using valid assessments, your odds are not like flipping a coin, or even rolling dice. They may be more like the chances of catching a black cat in a dark room.