Monday, July 26, 2010

Skills for Work vs Skills for Getting Hired

Given the amount of attention my previous security career post gathered (“A Myth ….”), it is time for a new one. Some of it is inspired by Source Boston 2010 mentoring panel, a gift that just keeps on giving (BTW, I signed up as a mentor with that new project, InfoSecMentors).

So, let’s talk about security skills that you can prove, skills that you need for a job and skills that will pass HR filters. It shocks me – to put it mildly – that these three are often completely different – and not even overlapping.

Which ones do you need to develop? Should you spend time writing papers, hacking code or reading up on 10 domains of “see-bee-kay”? Should you get good at something that will not be immediately obvious to everybody (like reversing malware) or spent time doing something visible (like writing papers about malware without having first-hand knowledge of how it works)? Should you choose sexy esoteric area of security, get really good at it – and then notice that nobody wants to hire you for that – with the possible exception of a Russian crime syndicate? :-)

While it is extremely tempting to bark “All of them!” and stop right there, the reality seems more complex to me, as it almost always is.

  • Skills that help pass HR filters (and especially certifications like “see-sssss-ph”) sure seem important as you won’t even have a chance to get to using your other skills aka be hired – unless you are a master-ninja-networker! By the way, buzzword - loading your resume is not about skills - it is about a socially acceptable form of lying: TCP/IP, UDP, ICMP, BGP, IDS, IPS, W3C, CIFS, WAF, DLP, GRC, SIEM, NAC, IAM, SNMP, SMTP, POP3, HTTP, NASL, IPv6 … ASS :-)
  • Skills that will help you do the job obviously vary depending on what job you have in mind. For most entry- and mid-level security roles, these skills are technical (sorry, Mssrs Security Policy Writers). From log analysis to IPS tuning to firewall management to web application scanning, the range is broad and you need to choose.  You can pick an area and then go really deep; however, it is worthwhile to try not to pick “typewriter repair” as an area of specialization :-) Fortunately, since none of the security problems we ever faced have been solved yet, choosing wrong is very hard. If you are still lost, pick application security or pentesting. These are not going away – EVER!
  • Skills that are easy to prove - typically via a multiple choice test - is another interesting set: some technical skills (such as knowledge about what is in TCP/IP header) are easy to test, while others (such as an ability to do web app penetration testing) are extremely hard to validate. I guess social engineering is an ultimate “unprovable” skill, while knowledge about how to configure a Cisco router is easier to prove. BTW, I’ve met some “Cisco Gear Master Magicians” whose skills bordered on divine – they can literally get that box to do anything.

And if I were to give some advice on this that I wish I received when I started in security, I’d say focus your energies like this:

  1. Put most of you energy in developing skills that will be most useful at work – work you do at your current job or the one you dream about (aka your next job :-)) As I said above, it is more likely that these skills are technical.
  2. However, balance the time you spent practicing technical skills that are simply fun for you with the ones that are easy to prove to potential employees. Let’s call them “visible skills.”
  3. Severely limit the time you spent on developing skills just to pass HR filters – instead get better at networking! Darn, even Twitter skills are better than practicing your daily laps in alphabet soup like the mess above.

To figure out that point, I once asked my wise mentor “Why do you still run /bin/bash, awk around and install Fedora, after you wrote three books, sold a company, gave a dozen keynote speeches and run a profitable consulting business for many years?” He – wisely, of course – said: “So that I can be a sysadmin if shit hits the fan.” This line is still stuck in my head after many years!!

Otherwise, you risk being of those types who respond to an ad for “firewall admin, must have CISSP” and end up crashing the network, which is kinda sad. For example, for many years I’ve had this bizarre unconscious skepticism towards people whose main skill is to write security policy. Writing this post cleared my head as to why: a well-written security policy does EXACTLY nothing for security … unless it is implemented.

Finally, some folks reading this will say – “screw the skills, I just want to be an expensive loudmouth for hire.” OK. There are indeed a few who rose to such noble occupation… First, you have to slave away for many years doing something else – and then hope that eventually people will want to pay to listen to your rants. Second, you can join Gartner, still slave away for a few years – and then maybe people will pay for your “loudmouthery.” In both cases, you’d still need some “+5” to Luck :-) And then maybe you can be “a mercenary loudmouth.”

But this is likely a subject of another post.

Possibly related posts:

Dr Anton Chuvakin