Response Time Monitoring
A site that is technically up but slow is still losing you customers.
Every website monitor tracks response time automatically, with an optional slow-response alert. For when that's all you actually want, Kawze also has its own lightweight monitor type - just a URL, an interval, and a response-time graph, with no up/down status or outage handling to configure at all.
How it works
- Every check Kawze runs, on any HTTP monitor, records the response time automatically - no separate monitor needed for that alone.
- Or add a dedicated response-time monitor: just a URL and an interval, purpose-built for tracking a trend rather than an outage.
- Set an optional slow-response threshold, on either monitor type, to get alerted separately from a full down/up incident.
- Response times are charted over time so you can see trends, not just a single data point.
Why it matters
- Surfaces gradual performance degradation before it becomes downtime.
- Historical response-time charts help correlate slowdowns with deploys or traffic spikes.
- The dedicated monitor type skips outage/notification bookkeeping entirely - just a graph, until you opt into alerting.
Frequently asked questions
Do I need a separate monitor for response time tracking?
No. Response time is recorded on every check of a regular website monitor. If you want a monitor that is only about response time - no up/down status at all - Kawze also has a dedicated response-time monitor type for that.
Does a slow response count as downtime?
Only if you set a threshold and it is crossed. Otherwise, response time is tracked and charted without affecting your uptime percentage.
What is a good website response time?
As a rough guide, most users notice a delay above 1-2 seconds and APIs are usually expected to respond in a few hundred milliseconds. What counts as "good" depends on your endpoint, so response time monitoring is most useful for tracking your own baseline and catching deviations from it.
Can I set a different response time threshold for each monitor?
Yes - the slow-response threshold is set per monitor, so a heavy dashboard page and a lightweight health-check endpoint can each have their own realistic threshold.
This capability lives on Kawze's timings monitor - see that page for the full feature list and pricing for that uptime monitor type.