Predictive tools in Talent Management (e.g., Recruiting) – proceed with caution!

Fact: Pure scientific method actually includes testing the hypothesis -- and the opposite of the hypothesis.

Opinion: I believe this purist approach to science and predictive tools unfortunately evades many modern-day “salaried scientists” … including some I-O (Industrial – Organizational) Psychologists who develop assessment tests to predict job success or flag questionable job candidates … sometimes only looking to confirm what they believe to be true. This is probably a function of the intense pressure to get verifiable "scientific" results more quickly in order to demonstrate business value to internal / external clients.

Many recruiting experts would agree that “false negatives” (rejecting a job candidate that would have been a great contributor) are much more harmful to an organization than “false positives” (hiring a job candidate that turns out to be a bad hire). This is generally the case because – as Bill Gates maintains – losing a potential star developer to a direct competitor can be the equivalent of losing $1 billion over the career of that developer. In contrast, most poor hiring decisions are usually addressed / ameliorated within 3-6 months; so on average, they are rarely costing an organization over $40-50,000 for the average professional. So – potentially, a “false negative” can be 20,000 times more costly than a “false positive” --- ok, maybe worst case.

In this context, perhaps we should be concerned that many assessment tests given to job candidates include this particular item to generally screen them out:

• A previous “job-hopping” pattern is often used to predict which job candidates would likely not be an ideal hire, even though the opposite may well be true for certain positions or job situations. For example, a business development exec that changes jobs every 2-3 years probably has a considerably bigger network of contacts to call on … and likely even a broader selling skills repertoire, than a sales executive who has been with one or two organizations over a long sales career. Moreover, a job-hopping sales exec may have been so good that their Sales Comp plan did not adequately reward them, driving them elsewhere.

• The same job-hopping disqualifier or “yellow flag” as a predictive tool also runs counter to the notion that people who have worked in a variety of organizations are exposed to many different ways of doing things, including a broader range of best industry practices.

Bottom line --- predictive tools can be very powerful and useful in Talent Management (e.g., Recruiting), but caution should be exercised in the form of not applying the same conclusions across all types of roles, candidate / manager behavioral profiles and work situations.