A while back, I published “Four Interview Questions I Have Asked Network Engineering Candidates,” a popular piece overall. I recently had an e-mail discussion with a friend looking for some more thoughts on how best to interview a candidate, so I thought I’d do this second blog post summarizing my answers.
I’m going to assume you read the other article first. I’m jumping in with more questions that follow on.
- What sort of job aspects keep you engaged? A problem for many working in IT is that we bore quickly. Resumes showing job changes every 2-3 years are not uncommon. Because of this, prospective employers must consider that they might be hiring an employee for a temporary period, i.e. a full-time employee that for practical purposes functions as a consultant or staff-for-hire. The strategy then is to figure out if it’s reasonable to expect that the prospect will stick around for at least 2 years, and more if things happen to be going well. To determine that, you have to figure out what the candidate likes. If the candidate likes projects and you have projects scheduled, well that will work out well. If the candidate likes to work independently and you need a self-starter, there’s a good match. If the candidate likes money but your budget is tight, that might not be a great fit.
- Describe your favorite manager you’ve ever worked for. This is a personality question. You want to discover what the candidate values in a manager. You can infer a lot about the candidate in their positive review. Maybe they like a manager that’s responsive to questions. Or standoffish. Or flexible. Or detail-oriented. Or invites them out for beer after work once in a while. Or offers frequent feedback. Whatever the answers are, you can gauge how the candidate will work with you (if you’re the manager) or with other leaders in the organization. Conversely, a bad question to ask is “tell me about the worst manager you ever had.” That puts the candidate in the awkward position of having to be negative in the interview, something that conscientious candidates will want to avoid. Getting an honest, frank answer to the “worst ever manager” question will be difficult, and you can discern about as much from the “best manager ever” question anyway.
- How would you solve this ethical dilemma? Tricky situations come up in IT. IT staff often have access to sensitive information, as well as the ability to obfuscate their tracks. In addition, IT jobs can be stressful and under appreciated, leading to staff that are tired, frustrated, and even angry. Some IT folks act as consultants, giving them privileged access to their clients’ business processes and data. Some businesses deal with information that is confidential and/or governed by regulations. Therefore, it may be appropriate to put a candidate into a hypothetical ethical dilemma, and ask them how they would handle it. This is one of those situations where the candidate might think they know the right answer no matter what they’d do in real life, but it’s still a discussion worth having. Would the candidate accept excessive gifts or kickbacks from a vendor? Would the candidate reset passwords outside of the normal process? What would the candidate do upon discovering a fellow employee had been sniffing financial data off of the wire?
- Sometimes not everything interesting or relevant is on their resume. Therefore, you might have to probe a bit. I have seen resumes or CVs where every protocol the candidate could think of was listed. I’ve also seen resumes where general skills were listed, but not specific networking technologies. In those cases, you will need to come up with your own list of technologies the candidate must have experience with to be successful, and ask open-ended questions. For example, I’d much rather ask, “OSPF or EIGRP?” just to get a discussion started than “What’s your favorite way to originate an OSPF default route?” Discussions are more enlightening than straightforward answers to narrowly scoped questions. You can always probe with more detailed questions later if you want.
- Consider asking a bogus question. By this, I mean that you should ask a question that’s purposely phrased with incorrect technical information, just to see how the candidate handles it. Admittedly, this is a bit diabolical. Assuming the candidate catches the error, they’ll probably want to be polite and gloss over or ignore it. Even so, press the issue. The concern here isn’t the question. The concern is to find out how the candidate handles correcting fellow team members. Are they polite? Rude? Condescending? Do they see the error as a mentoring opportunity? How do they react if you get defensive, insisting you are right and they are wrong? Some questions along these lines would be, “What is the default TTL value of an Ethernet frame?” (There is no TTL field in the Ethernet frame format.) “Explain how to get around the 2K limitation of 802.1q VLAN tags.” (802.1q VLAN tags are limited to 4,096 possible values, not 2K.) “How is the designated router elected in Cisco OSPF point-to-point networks?” (They aren’t. DRs and BDRs only show up on multiaccess networks like broadcast segments.)
All of these questions follow along with the central theme I presented in the first article, that of getting inside the candidate’s head. Technical chops are important, but personality and compatibility with the existing group are similarly important.