
MTTD vs. MTTR: What They Really Tell You About Your SOC
Mean Time to Detect (MTTD) answers one question: how long between when something malicious happens on your network and when your SOC notices it. Mean Time to Respond (MTTR) answers a separate one: once it’s noticed, how long until it’s contained. Contractors evaluating SOC providers often treat these as a single combined number. They shouldn’t. A provider can be fast at one and slow at the other, and the gap tells you where the actual risk sits. Our earlier post on what a managed Security Operations Center does touches on both metrics as supporting evidence of SOC performance; this post is the deeper explainer on what they mean and how to benchmark them.
Why the Two Numbers Move Independently
A SOC can have excellent detection tooling (SIEM correlation, behavioral analytics, threat intel feeds) and still have a slow MTTR if the response process depends on a single analyst working through a manual escalation chain. Conversely, a SOC with a tight, well-rehearsed response playbook won’t help you if the detection layer misses the signal in the first place.
This is also where staffing model matters more than tooling. A well-instrumented SIEM firing an alert at 3 a.m. doesn’t shorten MTTR if there’s no one watching the queue until 8 a.m. The alert sat there, detected, unaddressed, which on paper looks like fast detection and terrible response, but functionally is a 24×7 coverage gap wearing a metrics costume. Our guide comparing in-house vs. managed SOC staffing models gets into exactly how staffing depth changes both numbers in practice, not just in theory.
What Reasonable Benchmarks Look Like
Industry ranges vary by sector and control maturity, but as a general orientation, mature managed SOC environments target detection in minutes rather than hours, and containment of confirmed incidents within the same shift they were detected, not the next business day. IBM’s Cost of a Data Breach Report has repeatedly found that breaches identified and contained faster cost organizations substantially less than those that linger. NIST’s guidance on incident handling, referenced throughout the broader NIST SP 800-171 control set that underlies CMMC Level 2, frames detection and response as a continuous lifecycle rather than two isolated events: a slow MTTD means the response phase inherits a wider blast radius, and a slow MTTR means detection speed gets less credit than it deserves.
If your current numbers are measured in hours for detection or days for response, that’s usually a staffing or process gap, not a tooling gap. The number worth asking your provider for isn’t just an average. It’s the after-hours average specifically. A SOC that reports a strong blended MTTD/MTTR can still be masking a business-hours-only performance profile if nights and weekends are pulling the number in the wrong direction and getting averaged out.
Where 24×7 Coverage Moves Both Numbers
Continuous coverage compresses MTTD because there’s no window where alerts queue unwatched. It compresses MTTR because containment decisions don’t wait for someone to come online. They happen in the moment, by an analyst who’s already looking at the telemetry. The two metrics reinforce each other under a true 24×7 model in a way they can’t under a business-hours-plus-on-call one. We break down exactly how that staffing math works, and what it costs to close the gap, in our whitepaper, The SOC 24×7 Imperative. For a broader look at how continuous monitoring and layered detection work together at the network level, our comparison of NOC vs. SOC functions is a useful companion read.
How to Track and Question These Numbers
Tracking MTTD and MTTR well requires more than pulling an average from a SIEM dashboard at the end of the month. A useful tracking approach separates incidents by severity tier, separates business-hours from after-hours occurrences, and tracks trend lines over time rather than single snapshots. It’s also worth tracking a secondary metric alongside the two headline ones: Mean Time to Acknowledge (MTTA), how long between an alert firing and a human actually looking at it, which is often the hidden driver behind a sluggish MTTD or MTTR.
Beyond the after-hours breakdown already mentioned, ask your provider how MTTD and MTTR are actually measured (from what starting timestamp, verified against what ground truth, and reported over what window) and whether the numbers are tracked per-analyst or only at the team level. Aggregate numbers can mask a wide spread that matters most on overnight shifts, where staffing depth is often thinnest.
Metrics defined loosely enough can be made to look good regardless of actual performance, which is exactly why a specific, auditable definition matters as much as the number itself. A provider or internal team that can hand you a chart showing MTTD holding steady across day and night shifts over the past two quarters is demonstrating something meaningfully different from one that can only offer a single number for the whole period. That chart, more than any sales claim, is the artifact worth asking for before signing anything.
Want a benchmark conversation specific to your environment? Reach VSO at (888) 805-0510 or sales@vso-inc.com, or download our whitepaper, The SOC 24×7 Imperative, for the full staffing-and-cost framework.
FAQ
What’s a good MTTD for a managed SOC?
Mature environments generally target detection in minutes for high-severity activity. If your provider is quoting hours as a standard, ask whether that number includes off-hours incidents.
Can MTTR be fast even if MTTD is slow?
Technically, yes, but a slow MTTD usually means the attacker had more time to establish persistence before anyone responded, which tends to make the actual response harder even if the clock on it looks fine.
Should MTTD and MTTR be measured separately for business hours vs. after hours?
Yes. A blended average can hide a real coverage gap. Ask any SOC provider for the after-hours breakdown specifically.
What is MTTA and how does it relate to MTTD and MTTR?
Mean Time to Acknowledge measures how long an alert sits before a human looks at it. A slow MTTA often explains a sluggish MTTD or MTTR even when the underlying investigation, once started, moves quickly.
Should these metrics be tracked per-analyst or only as a team average?
Both are useful, but a team-only average can hide a wide spread in individual performance, which matters most on overnight shifts where a single analyst’s numbers can dominate the reported average.





