Website downtime tracker that reaches the right person
UptimePilot checks your sites, APIs and servers every minute from two regions, confirms a failure before it alerts anyone, and then tells the right person on the channel they actually read. If nobody responds, it escalates. When the site recovers, it says so, and every incident lands in a log with start, end, duration and cause.
Website downtime monitoring alerts on every channel your team uses
Different people need different channels. The on-call engineer wants a text message at night. The team wants a message in the channel where they already work. Your own tools want a machine readable event. UptimePilot sends all of them, and every plan includes every channel.
- Email. Subject line with monitor name, state and cause, so it can be filtered and read from a phone lock screen.
- SMS. Short text with monitor name and cause. Each plan includes a monthly SMS allowance (30 on Starter, 150 on Team, 500 on Business, 1,500 on Enterprise). When the allowance is used up, alerts continue on the other channels until the next month starts.
- Slack. A message in the channel you pick, with the state, cause and a link to the incident.
- Microsoft Teams. The same alert in a Teams channel.
- Webhook. A JSON request to your URL on down, recovery and acknowledgement, with monitor id, state, cause, region results and timestamps, so you can open tickets, post to a chat bot or trigger a script.
You choose channels per monitor or per group. A checkout page can go to SMS and Slack, a marketing blog only to email.
A downtime monitor that does not cry wolf
An alert you stop trusting is worse than none. Most false alarms come from one network path: a lost packet, a route that flaps, a probe that is briefly overloaded. UptimePilot marks a monitor down only after a second region confirms the failure. You can also require a number of failed checks in a row before an alert (two is a good default for a 1 minute interval) and set a timeout that matches the page (10 seconds for a heavy page, 5 for an API). The result is fewer alerts, and the alerts you get are real.
Alert Slack #ops and email the on-call engineer when 2 regions report down. Escalate by SMS after 5 minutes.
Escalation when an alert is not acknowledged
An alert sent at 3:00 to someone asleep is not a response. Escalation rules say what happens next. If the alert is not acknowledged within N minutes, UptimePilot alerts the next person or channel, and the next after that.
- Step 1, at once. Slack channel and email to the on-call engineer.
- Step 2, after 5 minutes without acknowledgement. SMS to the on-call engineer.
- Step 3, after 15 minutes. SMS to the engineering lead.
Acknowledging stops the escalation for that incident and shows the team who is on it. Rules can be different per group, so client sites of an agency can escalate to the account manager instead of an engineer.
Maintenance windows and recovery alerts
Deploys, migrations and server patches cause expected downtime. A maintenance window pauses alerts for the monitors you select, for the time you set, once or on a repeating schedule. Checks keep running, so you still see what happened, but nobody is paged, and the time is marked as maintenance in reports instead of downtime. If you publish a status page, the scheduled maintenance is shown there too.
When a confirmed outage ends, UptimePilot sends a recovery alert on the same channels with the total duration. Your team knows it can stand down, and the people who were escalated to do not need to ask.
An incident log you can use in a postmortem
Every confirmed outage becomes an incident with the details you need later.
- Start time, end time and duration, to the second.
- Cause, such as status code 503, timeout after 10 seconds, connection refused, keyword missing, certificate expired or heartbeat not received.
- Results from each region at the time of failure.
- Who was alerted, on which channel, when it was acknowledged and by whom.
The log feeds uptime percent and the incident list in uptime monitoring service reports, including the monthly SLA PDF on Team and above. A month with 99.9 percent uptime allows about 43.8 minutes of downtime, and the log shows exactly where those minutes went.
What the website downtime tracker watches
Alerts work the same for every check type: HTTP(S) status and response time, keyword on page, API checks with assertions, ping, TCP ports, SSL certificate, domain expiry and heartbeats from scheduled jobs. Read more on website monitoring, api monitoring, server monitoring tools and cron job monitoring.
Which plan fits your alerting
Email, Slack, Microsoft Teams and webhook alerts, escalation, maintenance windows and multi-region confirmation are in every plan. The plans differ by monitors, interval, SMS allowance and team size.
- Starter, $49 a month or $24 a month billed yearly ($288 a year). 50 monitors, 1 minute checks, 30 SMS a month, 3 team members.
- Team, $99 a month or $49 a month billed yearly ($588 a year). 200 monitors, 30 second checks, 150 SMS a month, 10 team members.
- Business, $249 a month or $124 a month billed yearly ($1,488 a year). 1,000 monitors, 500 SMS a month, 25 team members.
- Enterprise, $599 a month or $299 a month billed yearly ($3,588 a year). 3,000 monitors, 1,500 SMS a month, 100 team members, audit log and SSO (SAML).
Compare all limits on uptime monitoring pricing.
01
How quickly do downtime alerts arrive?
Checks run every minute on Starter and every 30 seconds on Team and above. The moment the second region confirms a failure, the alert goes out on the channels you chose.
02
What happens when my SMS allowance runs out?
SMS alerts stop until the next month begins. Email, Slack, Microsoft Teams and webhook alerts keep working without a limit, so no outage goes unreported.
03
Can different monitors alert different people?
Yes. Channels and escalation rules are set per monitor or per group, so each client, project or team gets its own routing.
04
Does maintenance count as downtime in reports?
No. Time inside a maintenance window is marked as maintenance and is kept out of the uptime percent, while the checks still record what happened.
05
What does the webhook send?
A JSON request on down, recovery and acknowledgement with monitor id, name, state, cause, region results and timestamps.
Know about downtime before your customers do
Add your first monitor, choose who gets alerted and how, and set an escalation step.