Job Interview Strategies (Part III)

[ This is part 3 in a 3-part series; if you haven't done so, you should start with part 1 and part 2 ]

I recently had a job audition. It was for a programming job and yes, I did say "audition'. This was the most interesting (and unique) interview / pre-screening technique I've seen to date. I approve wholeheartedly.

The job description was unusual:

MUST take a test before the interview

Have you developed with Perl, Javascript and XML in a UNIX environment? Are you interested in working at a dynamic company with leading edge technology? Would you spend 30-40 hours on a take home test, before the interview? If you are this ambitions, please send a resume. Qualified candidates will receive further instructions. This will change your life....

While I certainly didn't expect my life to change, I was, nonetheless, intrigued. I sent my resume; I got a call back; we talked. The recruiter gave me his take on the "test" idea. The test serves two purposes. First, it provides the client company with a much more realistic picture of the applicant's abilities and skills.

Second, by getting people to commit 30 to 40 hours on a test, the company has a much better idea of the "sticking power" and motivation of the candidates. Candidates self-select before the interview. Those who aren't really qualified, aren't motivated, or simply aren't really interested, drop out before finishing the test.

When I first described all of this to a friend of mine, he was cynically unimpressed. I think his next words were something along the lines of "Eh. How high can you jump?" We're so hardened to the concept of the troll at the gate playing buzzword bingo with no real clues that he simply assumed this was another 'typical ill-considered pre-screening tactic. I had to reassure him that the test was, in fact, coming from the client company and would be evaluated by the client company, not the recruiting agency.

I also had to assure him that I thought the test was a good idea. (He still wasn't convinced). In any case, I signed and returned an NDA to the recruiter and was sent "the test".

It wasn't at all what I was expecting. I was expecting questions and multiple short problems. What I got was in the form of a practicum exam, described thusly:

The following exercise provides prospective application engineering partners with a practical example of the work required to build a simple IVR solution within the context of [Company X's] value-added development process. The exercise is designed to be completed in one weekend by one or two engineers possessing strong command of VoiceXML, JavaScript, and Perl. Certain aspects of the exercise are intentionally under specified; engineers should make appropriate implementation assumptions and document these clearly.

When I described the actual "test" to my friend, his response was even more cynical. "So", he said "they're getting you to do their work for them for free. Is this how they write their code for customers?" I don't think so; I believe this is a standard exercise that Company X keeps on hand for new OEMs, customers who don't want to pay the company consultants to write code for them, and, well, interview candidates. Why not?

I was told, by the recruiter, that the test was expected to take 30 or 40 hours "in your spare time". It took me 46 hours over the period of a week. In the process I learned a little bit more about JavaScript (which I find I still don't much care for) and a lot about VoiceXML (which differs from XML primarily in that VoiceXML is a programming language; I'm still not sure if I approve of that idea! :)

I learned a lot about voice recognition systems and voice mazes. I uncovered a few bugs in Company X's development tools and documentation (as well as a few in the test itself ;-). I have some new buzzwords I can add to my resume.

I don't know if I'll be asked to come in for an interview (although I think I should be, after committing my time and energy; commitment at that level should be a two-way street!). Also, after working through the exercise, I realized that I'm ambivalent about whether I would want to do that sort of work fulltime for the next 3 to 6 months.

Nevertheless, I am very impressed. This is what an interview should be.

There are two major goals in an interview. The candidate's goal is to learn whether he or she wants to work for the company being interviewed. The company (in the person of a manager or two and some potential co-workers) wants to discover if the candidate is qualified to join their team — qualified both in terms of skills and personality.

Both sides prefer to make their determination as son as possible, without wasting too much of their valuable time. neither side wants to make the wrong decision. In fact, I've heard hiring managers admit that they would rather mistakenly dismiss a promising candidate than make a bad decision to hire someone that turns out to be the wrong fit.

The pre-screening exercise I recently went through for Company X meets both goals admirably. It does an excellent job of weeding out candidates who are taking a shotgun approach to job hunting — sending their resumes to any job that looks even remotely possible within a 100-mile radius. Completion of the exercise tells the company about the applicant's skills in coding, in documentation, in testing. Completion also gives the company some evidence of the candidate's level of commitment and motivation.

The same exercise serves to meet the goals of the candidate as well. After spending a week immersed in this project, I have a very clear idea of what Company X does! I've had an excellent introduction to VoiceXML and a practical example of what my days would be like if I were to be offered (and were to accept) this job! I've never before felt this familiar with how it would feel to work for a company — not without having first worked at the company for a few weeks!

I realize that most companies don't have exercises like this one "just lying around", ready to hand to candidates. However, I think such exercises are an excellent idea. More companies should create exercises of this form. I believe a practical programming project is far more accurate than a half hour Q&A session when it comes to determining a technical candidates qualifications for a technical position.