How to hire for “fit”

I have the pleasure of chairing a CEO Peer-to-Peer group and one of the recurring topics is about fit.  
How do we ensure that new employees we bring into our company are a good fit for the company and the role for which we hired them?

We are not talking about their resume or their skill level.  We are talking about their attitudes, their values, and their behaviours.  The things that really determine whether or not they "fit" and will be able to contribute over the long-term to the success of the organization.

And no, we are NOT talking about making sure everyone in the company is just like us and did exactly what we wanted.  The Stepford Wives  has shown us the error in that thinking.  We want diversity of thought and we want respectful differences and debate.  But we all know that that elusive "fit" is real so we wanted to get better at figuring it out and doing our recruiting in a way that made sense.

So we brought in a guest speaker, Neil MacGregor, from Plum ( Plum website ) to help us work through some thinking on this topic.  Plum provides an online assessment tool which is used by both management and job applicants as a way to check not just what applicants can do but how they will work together with management.

We started off by all agreeing that:

• an interview (series) can determine the relevant skills someone has (relatively) well
• determining an applicant's "fit" (attitude, values, ...) with the organization
       and the role is much more difficult ( and not usually done well )
• getting the "fit" wrong will likely end up with a dysfunctional employee or a termination procedure
      ( 89% of employee turnover is due to attitude misalignment,
        11% due to skill deficiency ... Mark Murphy in Hiring for Attitude)
• getting the "fit" wrong is very expensive for an organization
      ( the all-in MINIMUM cost to a company to let someone go
         is 30% of their annual salary  (US Department of Labour)
         and see the Infographic from Mindflash on the "Staggering costs of a bad hire" )
• you need more than 90 days to properly assess if the http://www.plum.iohttp://Fast Company aritcle on costsperson will work out in the role

But is it really that bad if we mess up for just one person in one role?  

A bad hire can pull down the productivity of the rest of the team.  According to Maximilien Ringelmann (1861–1931) and a conclusion named after him called the Ringelmann effect, having group members work together on a task (e.g., pulling a rope) actually results in significantly less effort from each individual than when individual members are acting alone.  Group members tend to rely on their co-workers to furnish the desired effort required for a communal task.  In spite of our best hopes, in a characteristic called "Social Loafing", the efforts of the best of the group will decrease to match the efforts of the worst of the group instead of the other way around.  There are techniques to minimize "Social Loafing" but the point is, a bad hire can pull down the productivity of the rest of the team.

So how do we test for attitude?  And how do we determine what attitude is right for the company and role?

Behaviour is a function of Attitude.
Attitude is a function of Values.

So to determine attitude, look to values.

Psychometric testing (to uncover values) is a complex topic and many commercial assessment profiling tools ask questions in a way that makes the underlying metric transparent, thereby encouraging participants to second-guess the questions and answer what they think they should answer, instead of what they truly feel.  This creates invalid results, making some assessment tools part of the problem, not part of the solution.

Personality profiling tools like Myers Briggs do provide ways to talk about our differences and similarities.  That helps individuals and teams create better understanding amongst disparate "types" and can be used to foster better working and personal relationships.  But their scientific validity in determining fit for a particular role is often questioned - see  Washington Post article on Myers Briggs for an article on this from the Washington Post.

Additionally, most profiling is done late in the recruitment process, after the short list has been created, to determine who, from the short list, should be considered for further investigation.  Since "fit" is so important, it is recommended to reverse the process and use "fit" as the first filter, creating a short list that only includes candidates who "fit".  Then use skills as the second filter.  This is the Plum approach and requires that the profiling tool pricing model and method be set up to allow low cost, high volume assessments.

So you are now down to a few, short-listed candidates.  How do you select the best one?

Testing for intelligence is one method.  It is no surprise to learn that intelligent people are more likely to succeed than non-intelligent people (yes, someone paid to have this research done).  But intelligence has many facets; neurologists are still developing frameworks to better understand and work with the different kinds of intelligence.  We have all heard of IQ and EQ.  Another framework talks about Fluid and Crystallized intelligence as described in the Scientific American article at  Scientific American article on intelligence types .   A New York Times article at NY Times article on intelligence types can help you get a better sense of the Intelligence types .  Yet other theories have come along (and some gone away) so for now, determining if a person is "intelligent" is useful but incomplete.

One of the best methods is a "job tryout", where you have each short-listed candidate do something they would do in the role and you assess them on their results.  In addition to getting a real sense of who can do the job well, you allow people who may not be as good an interviewer (eg. introverts don't like talking about themselves) to shine through (eg. introverts may produce the best results and are comfortable talking about their project).

Another, related method is to give each candidate a problem to solve.  Problem-solving skills are key indicators of future success in a role, so if the candidate can do a good job solving a problem relevant to the role, odds are they will perform well in the role.

One caveat: Past success is NOT a good predictor of future success in a different environment.  Neil quoted a statistic that surprised some: only 10%-15% of people can succeed anywhere, in any environment.  85% of people who can succeed can do so only in certain environmental conditions.  So when doing a job tryout or problem-solving test, try to duplicate the environment in which the candidate will be working; this will give you the best indicator of future success for the candidate.

So, to hire for "fit":
- Use an appropriate assessment tool on all applicants as a first filter
- Create the short list of skilled candidates from those who made it through the first filter
- Use job tryouts or problem-solving tests with the short-listed candidates

Thank you to Neil MacGregor for guiding us in this discussion.
Note that I am not affiliated with nor have I made any investment in Plum.  I do like what I see in Plum, but I encourage you to do your own research and choose the assessment tool you think serves your purpose.