It's just human nature. Incentives drive behavior. And such is the problem of doing security. What's the incentive for doing security well? How do you even know if you're doing security well? If you beat back the attackers, what happens? Nothing. Actually, you don't get kicked in the teeth that day. So there's that. If you miss something and it results in data loss or downtime, get ready to take your lumps. It's a thankless job. And your incentive to excel is keeping your job -- which on most days doesn't feel like an incentive, right?
So how do we solve this problem? We have to define proper incentives for a security team. It certainly can't be stuff that doesn't impact the business. So if your MBO is based on shortening patch windows or full AV coverage, that's a huge fail. How about a bonus for achieving PCI compliance? Yeah, since PCI is such a high bar for security, you should definitely be incenting the security team to achieve mediocrity.
Likewise, if your success criteria (and associate bonus plan) hinges on having no data losses or breaches, you are being set up for failure. The fact is incidents happen. Whether we like it or not. You could do everything right and still get pwned. You want an incentive plan that doesn't ignore failure, but also puts a bulk of the incentive on things within your control.
Given that fact, what would be most impactful to the business? Maybe time to remediate an incident? Set a baseline and set objectives for improvement. That would impact the business, no? Of course, that can be a little squishy, but you have to start somewhere. How about something that prevents train wrecks before the train leaves the station? By that I'm referring to application security. Maybe something like number of security defects identified and fixed before applications go to production. That's not exactly in your control, but it's pretty important to the business.
You can quibble all you want about whether your security awareness training helps secure your environment. Yet with rampant social engineering attacks, how can you not spend some time training the weakest link in the chain -- your users? How about tracking the results of your internal social engineering simulations and providing incentives for improvement in the rank and file?
And it's not just the security team who need incentives to achieve security objectives. What about the operations folks who drag their heels on a firewall change (or, even worse, make a mistake), causing a potentially exploitable situation? What is their incentive to take a chance the firewall change causes downtime by closing a critical port? And how do you expect developers to write secure code when their incentive is to ship working code on time and within the budget? Right, they won't. Incentives create alignment within organizations as well. Without that alignment, achieving security goals is mission impossible.
This sounds an awful lot like a metrics discussion -- and it is. We, as an industry, have done a poor job of standardizing on a set of success criteria that indicate acceptable security posture. Without that success criteria and accompanying metrics to drive that behavior, you can't really design incentive programs to get both the security team, operations, developers, and the employee base to do the right things.
By the way, it's not like I have a lot of answers to address this issue. In my varied travels, it's hard to find folks who are happy with the incentive programs for the security team. Incentives are either nonexistent or based on activity that doesn't really reflect security posture or help the business. Until we get a much better handle on these incentives, folks will continue to ignore the need to secure things. That means we'll continue to repeat Groundhog Day over and over again.
Although I guess we could go all Game of Thrones and parade the head of the latest phishing victim on a stick in the cafeteria. That would certainly get everyone's attention, no?