Mean Time to Detect

What Does It Mean?

Mean time to detect (or MTTD) refers to the time it takes from when a problem first emerges to the moment when it is detected by the right people or systems. The KPI tracks how effective IT departments and organizations are at avoiding long-term disruptions and uncovering issues that could become problematic.

A ServiceNow Weekly Analysis Dashboard

Why Does it Matter?

Bugs, errors, and malfunctions may start off innocuously, but they can quickly become serious issues the longer they’re left undetected and unresolved. MTTD presents a clear snapshot of how effectively your organization can detect these problems and start the resolution clock. Most crucially, undetected errors can cause an impact on your revenues — from service outages that violate your SLA terms to server downtime that can scare off customers — and harm your reputation. Tracking your MTTD also helps you create better systems to detect issues, be they automated (in the form of error logs and bug detectors) or manual (having team members monitor implementations and deployments).

How Do you Measure the KPI?

The mean time to detect formula is quite straightforward when you have the right data to start from. To find the MTTD, simply add all the incident detection times for the given team member or time period and divide by the number of incidents. For instance, if your total time to detection for January is 850 minutes, and there were 12 incidents reported, your mean time to detect would be 70.83 minutes.

What Data Sources Would You Use to Measure the KPI?

The most important data source when it comes to measuring your MTTD is your software or service’s automated error reports. While you may not detect them immediately, most applications will note an abnormal behavior immediately in logs, even though you may not see a report until later. Additionally, manual error reports and support claims filed by your clients or internal stakeholders are also important.

Give me an example…

Let’s say you’re getting a large number of error complaints in a short amount of time, and you can’t seem to stay on top of them. The problem may not lie in how quickly they’re being resolved once they’re detected, but in how long they take to be reported. In our IT dashboard example, tracking MTTD can tell you that the problem starts earlier, and it can also explain why your mean time to resolution is so high. More importantly, it can illustrate if the problem lies in your automated systems, or if your team is too slow to detect issues.

What Benchmarks/Indicators Should I Use?

- SLA benchmarks
- Total time to detect
- Total number of incidents

More IT KPIs Start Free Trial