Have you ever hired someone who had their CV listing every programming language and so many certs that when took an interview they stared blankly the time you asked them to fix something then-and-there right in front of you? It happens more often than people care to believe. What we need to really work out in terms of how best to test candidates’ IT skills in NZ essentially breaks down into one brutally simple question; how do you detect real skill vs someone who is just good at self-describing before you’ve already made a costly error?

Because of heavy competition and oversaturation in the market, IT hiring is riskier than many other roles. This is partly because a poor decision can be really expensive and, partly, because the gaps aren’t always obvious right away. A developer who does not know how to write clean code or a support tech struggling with basic networking can quietly consume the business for weeks before he is detected.

Most people start with solving coding challenges. Outline a problem, set a time limit, see how they think. Not easy, not perfect, but it is a fair way to view and assess logic per candidate end masse.

But take-home assignments are a whole different animal. Because no one is looking over their shoulder, you get a more honest glimpse of what someone really codes day to day versus how they perform when they know they’re being evaluated in real time.

Then you have live coding, where someone senior literally sits next to the candidate and watches them work out a problem IRL. That tells you more than if the code works. This indicates how someone thinks when stuck, and whether a hint truly helps with solving the problem or complicates things even further.

Pair programming takes that one step further as it puts the candidate with a real team member for many teams, this is definitely the best way to assess IT skills in NZ collaboratively. You can observe communication with technical ability together, which in truth provides greater meaning than either independently.

This brings us to whiteboard-style interviews which still persist, though mostly for checking one’s grasp of algorithms and system design. Think of them not as a quiz; more as a conversation and they work best. Let candidates consider trade-offs, rather than trying to remember a line-by-line answer that they half-recalled from some textbook.

For non-development roles, support, testing, or infrastructure, scenario-based cases are fundamentally more useful than pure coding tests. Give somebody a real mess, a failed server or network misconfiguration, and put them under some pressure on how they’d troubleshoot the situation.

Online assessment platforms have also become the choice for early screening, especially when a cumbersome list of applications by a business is looming. On completion of a coding task and technical multiple-choice questions, they can evaluate the responses automatically before even human involvement, thereby saving significant time.

What matters is to choose the right combination rather than one immaculate test. Such basic approaches steer most companies towards some kind of preliminary online screening, which is then undoubtedly followed by a subsequent live or case-based step, where the speed vs candidate volume battle is fought hard against the depth you pragmatically need to ensure confidence.

Seniority alters the way the game is played as well, and it should be how you best assess a candidate’s IT skills in NZ seldom appears uniform across levels of experience. For junior candidates you can have a take home task and then walk them through their solution. Yet given how much more is at stake, senior and above hires require a wider scope live code, architecture discussions, perhaps even a glimpse into their history of work.
Simulations, wherever feasible particularly for infrastructure are definitely worth including. Plop someone in a simulated environment representing the real thing, give it an intentional defect, and watch how they diagnose and repair the issues. In my experience, it shows real practical capability far better than any theory ever does.

This doesn’t even work if you ignore the soft stuff. That they matter is no longer a theory, and how someone explains their reasoning, how they react when a fix doesn’t work the first time; those things count for just as much once they’re actually on a team. A mind-meltingly awesome programmer who is so uncompromising that they cannot accept criticism or explain their decisions to a non-technical colleague causes friction that outweighs the pure skill.

Achieving the best results usually arises from both performing structured technical testing and a frank discussion of how they actually do their work. A test tells you what someone is capable of doing. The chat informs you more about whether they’re going to gel with the way things actually work day-to-day on your team.

Conclusion

Not one test that makes or breaks IT hiring by itself, and perhaps never will. Combine structured coding assessments with a real task, what we call scenario-based tasks, and a sincere chat for the most complete overall view. It is actually fitting the approach to seniority instead of running everyone through the same arbitrary board that leads to minimizing expensive mis-hires over time.