Maturity Is Not a Test You Pass
One of the first questions I get when talking about security maturity is some version of "What score do we need?" Leadership wants a number. The board wants a number. The team wants to know when they are done.
It is a fair instinct. Most of us grew up with tests, grades, and certifications that had a passing mark. Even certifications in our own industry work this way. Hit the number, get the cert, move on. It is natural to want security maturity to work the same way.
This unfortunately is not the way security maturity works. There is no test to pass and no perfect score to reach but, that’s okay. The goal of measuring security maturity is not to arrive somewhere. It is to keep getting better, continuously, forever.
Why There Is No Perfect Score
A perfect score implies a fixed target. Security does not have one. Everything that defines what "mature" looks like for your organization is constantly moving:
- The threat landscape changes. New techniques emerge, old ones get repurposed, and adversaries adapt to the defenses that worked last year.
- Your environment changes. You migrate to the cloud, adopt new SaaS tools, acquire a company, or retire a data center. Every change shifts your attack surface and your visibility.
- Your team changes. People join, leave, and grow. Knowledge walks out the door and new skills walk in.
- The frameworks change. Even the yardsticks we measure against, like MITRE ATT&CK, are revised and restructured on a regular basis.
A program that was in great shape a year ago can drift backward without anyone touching it, simply because the world around it moved. Maturity is not a place you get to. It is something you maintain.
SIEMs are a great example of this. On the surface they all accomplish the same thing more or less. However, when we start to look under the hood and deeper into features and query languages each SIEM platform becomes quite unique, so measuring your security maturity between platforms is usually an apples to oranges comparison. The core idea here of continuous improvement applies in different ways between the different platforms and it’s something I’ve seen a number of times in my years performing SIEM migrations.
What Maturity Models Are Good For
None of this means maturity models are useless. They are valuable when used for what they are: a shared vocabulary and a map of what good looks like across people, process, and technology. They help you see where you are, where your gaps are, and what a reasonable next step might be.
They become a problem when the score turns into the goal. A few common traps:
- Chasing the number. Teams optimize for the checklist instead of for real capability. A control gets marked "implemented" even though nobody has validated that it works.
- Treating the scale as linear. Moving from level 2 to level 3 in one area may matter far more than reaching level 4 in another, depending on your threat profile and organizational goals.
- Measuring once. A point-in-time assessment tells you where you were on the day of the assessment. It says nothing about whether you are improving.
- Comparing across organizations. Two companies with the same score can have very different risk, exposure, and priorities. A score is only meaningful against your own history.
Use the model as a compass, not a scoreboard.
Measure Direction, Not Position
If there is no finish line, the most useful question isn’t "Where are we?" but "Are we getting better?" That shifts measurement away from a single score and toward trends.
For a detection and response program, some trend metrics I find useful:
- Detection coverage against your threat profile, tracked over time rather than as a raw percentage of ATT&CK
- Validated coverage, meaning detections that have actually been tested and shown to fire, not just deployed
- Alert efficacy, the ratio of alerts that lead to meaningful action versus noise
- Mean time to detect and respond, and whether it is trending down
- Rule and log source health, including how quickly broken detections or silent data sources are caught and fixed
- Backlog movement, whether known gaps are being closed faster than new ones appear
None of these have a perfect value. A program with 60% validated coverage that was at 40% last quarter is in a healthier place than one sitting at 75% that has not changed in a year.
An example of a problem I’ve seen with this in the real world is a company having their detection coverage percentage stay the same, while ignoring onboarding new data sources. Sure, 75% coverage is great, but when it doesn’t accurately reflect how the organization’s attack surface has evolved it’s not really something to celebrate.
Building a Continuous Improvement Rhythm
Continuous improvement does not happen by accident. It needs a cadence. The approach I recommend is simple:
- Establish an honest baseline. Assess where you are today without trying to look good. An inflated baseline only hides the progress you make later.
- Set a small number of goals for the next period. Pick the improvements that matter most against your threat profile, not the ones that move a score fastest.
- Define the metrics that show progress on those goals. Keep them few, specific, and measurable.
- Review on a regular cadence. Quarterly works well for most teams. Look at what moved, what did not, and why.
- Adjust and repeat. Retire goals you have met, add new ones as the environment and threats change, and keep going.
The review is where the value is. It turns maturity from an annual report card into an ongoing conversation about what to improve next.It also changes how teams feel about measurement. When nobody expects a perfect score, a gap stops being a failure and becomes the next item on the roadmap.
Better Than Last Quarter
There is no test to pass and no perfect score waiting at the end. The organizations that do security well are not the ones with the highest number on an assessment. They are the ones that keep looking honestly at where they stand, pick the next most important thing to improve, and come back to check whether it worked.
That is the mindset behind how we work with customers at Rilevera. We help teams set a realistic baseline, choose goals that fit their threats, and review progress together on a regular cadence. If you want to talk about what continuous improvement could look like for your program, we would be glad to have that conversation.



