5 Secrets of a Mature Vulnerability Management Program from Costa Coffee and Priceline

The 5 Secrets of a Mature Vulnerability Management Program

This week HackerOne hosted a series of webinars that asked participants about how they rated their level of vulnerability management maturity. We asked how regularly they tested their digital assets, which assets they tested and how they prioritized them, and whether they had a way for friendly hackers to tell them about any vulnerabilities. At least 90% had some sort of prioritization process in place to deal with vulnerabilities, only 28% conducted continuous security testing, but 59% admitted to having no way for third party researchers to submit bugs. 

Of course, a company’s level of vulnerability management maturity usually aligns with how sophisticated and extensive their technology use is. A mature vulnerability manager will know what assets they have, assess them regularly, remediate any vulnerabilities and feed that information back into the SDLC - and recognize that it can’t be done without the help of third party researchers to close the gaps. 

During the webinars, we caught up with Matt Southworth, CISO of Priceline, and Matt Adams, Global Security Architect at Costa Coffee, to learn their 5 secrets to building a highly effective vulnerability management program. One that embraces uses a variety of tools, that communicates effectively, that covers all bases, can adapt to changing needs and, above all, can demonstrate success.


Matt A: “At Costa we use automated scanners find the ‘low hanging fruit’ and highlight those areas we need to spend more time on, along with static and dynamic testing during the product release timeline to detect common vulnerabilities, such as those found in the OWASP Top 10. After that, we hand things off to the bug bounty program to find the things we couldn’t find ourselves. We expect really high value results from bug bounty as these vulnerabilities will be things the automated scanners haven’t picked up, and typically take human creativity and/or ingenuity to exploit. For example, we had a report where a hacker found a registration endpoint for a portal, which, although it had been superficially removed, would have allowed them to create an administrator account on a customer facing website by manually building an API request to trigger the account creation process.”


Matt S: “The mistake we see most often is when there’s confusion over who owns the app that a vulnerability has been found in and it gets sent to the wrong team, slowing the whole process down. Having clear dashboards that show where you’re at, what the problems are and, most importantly, where those problems map back to and who the owner is, will help speed that remediation process up. The difference between vulnerability assessment and vulnerability management is that assessment is a process with input and output, whereas management is about being able to communicate clearly to your organization about risk tolerance. Once you can make your stakeholders care about the risk you should then over-communicate to demonstrate where you’re at and what the pain points are.”


Matt S:” Every organization has blind spots. Having the hacker community on the other side of the screen looking at those things you’ve missed means you can close those holes. Blind spots could be anything from an employee purchasing software you’re not aware of to low level vulnerabilities that when chained together could have a significant impact. The most common blind spot we find is regressions; when you think you’ve identified and solved an issue but it’s still there or returns after the fix. That’s where retesting comes in to provide an extra layer of security.”


Matt A: “We have different patching schedules for different devices and services in our estate, largely dictated by each vendor’s release schedule. We adhere to those patching cycles as much as possible, but, when products have more sensitive and complex releases, we will add in additional testing cycles. If we’ve seen a vulnerability exploited in the wild, or had a warning from an industry body about a vulnerability, then it’s important that we have the flexibility to be able to patch ahead of the planned schedule. You need to find the balance between maintaining the usual patch process and the disruption of ad-hoc patches.”


Matt S: “Categorising success can be as simple as measuring the time to detect vulnerabilities, the time to fix and how the team is hitting the SLAs to fix them. Having visibility of which vulnerabilities are popping up over and over means you can focus on where the problems are, and work towards educating those system owners on how to avoid them.”


To listen to the webinars in full, you can find them on demand here: https://www.hackerone.com/resources/on-demand-videos 

The Ultimate Guide to Managing Ethical and Security Risks in AI

AI Ebook