Everybody does interviews.  It's the most common selection tecchnique in recuitment.  However, a brief tour on the internet will readily refer to studies that show that the typical interview (more specifically - an unstructured interview - see below) is almost worthless as a predictor of long-term on the job performance. In summary, you have only slightly better odds of choosing the better of two employees following the typical half-hour interview as you have by flipping a coin.

Why do most interviews fail?

In our view the main problem is, that for all too many employers, performance at interview has a huge weighting in terms of who gets the job.  It becomes all about which candidate "performed the best" at interview. Few managers get proper interview training and many also don't interview that often. The result is that they very often do, what is referred to as the unstructured interview - "a get to know you chat to see if you will fit in".  Ring any bells?

Such interviews can end up being highly selective (and not for the right reasons).  They may easily end up being based around whatever assumptions (prejudices) the recruiter or hiring manager has.  They can be easily affected by the mood and chemistry between them and the candidate. And, they can also be affected by factors such as; physical appearance and tone of voice. Let alone the problem areas of; race, age, sex and disabilities.  Studies show, even the time of day can cause variance in terms of how an interview is conducted and its impact on candidate perception.

The reason most managers take the unstructured approach. Or the reason most talked about, is that the alternatives take a lot more effort up front. Carefully constructed structured interviews take much more time to prepare and to achieve best effect, they have to be delivered in a very similar manner to all candidates.

Some other issues, by way of example, that can result from relying on unstructured interviews:

  • Managers will, other things being equal, tend to hire people who remind them of themselves.  It just human nature.  The problem with this. Is that it will likely reduce diversity and favour candidates based on factors irrelevant to their ability to do the job at hand.
  • Context.  Unless you guard carefully again this. People naturally tend to ignore the context or situation when judging people by results. Studies confirm managers will rate candidates higher, even when it is clear that they faced easier circumstances than others they interviewed.

If you are looking to engage your managers and don't want to wallow in the negatives around what they may be doing now.  You could raise this. Most managers we have seen, will acknowledge from their experience that; 'some candidates just seem to interview better...'  The studies suggest managers really like it when a candidate 'interviews well'. The danger however, is that they may all too often, equate this with a likely better performance on the job.  When the reality is that there is, more than likely, no correlation at all. By way of an example; extroverts tend to do better in interviews. But for many, if not most jobs, extroversion is not what the job requires (i.e. it's irrelevant). This example is one we have found many managers can relate to and its helped set the scene when discussing bringing in new selection techniques, that may otherwise see a bit of resistance.


What works?

If we are talking Interviews.   Then it's carefully constructed 'structured interviews'. These are where questions are directly related to: measure-able skills, behaviours, competencies, or past experience.   The structured interview method we like what is known as 'Behavioural Competency' based interviewing.  It's a technique professional HR managers often learn to use. But in simple terms it comes down to providing a 'structured' process.  One where each candidate gets asked the same questions and where these questions are directly related towards factors or behaviours you are assessing for job performance.

If you are currently doing unstructured interviewing and want to do better.  Why not try the following:

  • Identify 2 or 3 main priorities or factors that you are looking for in your ideal candidate.  
  • Take these and any other items you have good reason to believe will indicate better on job performance and write them down as your 'Assessment Criteria'.  
  • Measure (score) each candidate on how well they achieved your 'Assessment Criteria'. Especially your 2 or 3 main priorities (see above).
  • Use a script for interviews. Write the questions down in advance. It's what the best professional interviewers do. You really don't need many questions. But stick to the ones you have, don't get distracted. And, write down a summary of the answers you get from them.
  • When a candidate talks about their experience or achievements. Consider their situation and weight their response according to how hard it was, or what they actually had to do when you compare it with the answers from others.
  • Lastly - do try and avoid any interruptions (they skew the interview) and remember the old adage: "you have 2 ears and one mouth - use them in that order".  Basically remember to listen, listen, listen...

Anything else?

It would be very remiss in a post about interviewing not to have brief discussion about the alternatives. Even with the best interviews they are by no means perfect. We suggest your selection process should also include other, more objective evidence e.g. school reports, prior performance metrics, letters of reference, recommendations, etc. They typically offer better odds than an interview (studies suggest they are better than a unstructured interview by between 60-75%). 

Other things you could use, include: screening and testing tools; asking relevant questions to everyone who applies at the time they apply i.e. in the application process; realistic job task simulations; and aptitude, verbal and numerical reasoning tests. All these items will deliver better odds of picking  better candidates than an interview. Whilst some of these things cost money and may need external help, there are also some simple things you can do that don't.  As an example, why not ask candidates to do something related to the job you expect them to do.  Or, if feasible look at their previous work (but do please check it was them who really did it...). 

In Conclusion

If you have only really relied on interviews in the past. We suggest having several parts to your selection process will make it much more robust and you will have a better chance of securing those candidates that do bring that little bit extra to your business. And, if you currently do unstructured type interviews.   Try the more structured process we laid out above.  Yes, its a bit more work up front.  But its based on good practice and we suggest the effort will be repaid many, many, times with better performing staff.  As an extra bonus, you might also find you end up interviewing less, as you won't need to replace as many...