Website monitoring service that confirms before it alerts
UptimePilot checks your website every minute, or every 30 seconds on Team and higher plans, from US East and EU Central. It verifies the status code, the response time and the content of the page, watches the SSL certificate and the domain registration alongside, and alerts the right person only after a second region confirms the problem.
What website monitoring catches
A site rarely goes down in one clean, obvious way. The server answers but returns a 502 from a broken upstream. The page loads but the product grid is empty because a database query timed out. A deploy redirects every URL to the login page. The certificate expired at midnight and browsers show a full screen warning. Each of these looks fine to a simple ping and costs real orders, signups or leads.
Website monitoring is the practice of requesting your pages on a schedule from outside your network, exactly as a visitor would, and checking that the answer is correct. Done well, it tells you about the problem minutes after it starts, not hours later from an unhappy customer.
- Server errors (5xx) and unexpected client errors (4xx) on key pages
- Slow responses that cross your threshold, even when the page eventually loads
- Missing content, such as an "Add to cart" button or a price that should be on the page
- Error text that should never appear, such as "Database error" or "Maintenance mode"
- Redirect loops and redirects to the wrong host
- Expiring SSL certificates and broken certificate chains
- Domains close to their registration expiry date
Website monitoring service with every check type included
| Check | What UptimePilot verifies | Typical use |
|---|---|---|
| HTTP(S) | Status code, total response time, redirect chain | Homepage, landing pages, login |
| Keyword present | A word or phrase appears in the response body | "Add to cart", product name, footer text |
| Keyword absent | A word or phrase does not appear | "Error", "Exception", "Out of stock" on a key product |
| API with assertions | Method, headers, body, JSON fields | Backend endpoints behind the site |
| SSL certificate | Days until expiry and chain validity | Every hostname you serve |
| Domain expiry | Registration expiry date | Every domain you own |
All check types come with every plan. You do not pay extra for keyword checks or for SSL monitoring. For endpoints that return JSON, see api monitoring. For certificate details, see ssl certificate monitoring.
How site availability monitoring works on UptimePilot
- 01 Enter a URL and choose the check type. For a store, add a keyword such as "Add to cart". For a SaaS app, monitor the login page and the status of the main API.
- 02 Choose the interval. Starter checks every minute. Team, Business and Enterprise check every 30 seconds.
- 03 Set thresholds. A response time limit (for instance 3 seconds), a timeout (for instance 10 seconds) and the expected status code or range.
- 04 Pick who gets alerted and how: email, SMS, Slack, Microsoft Teams or a webhook. Add an escalation step, so the next person hears about it if the first one does not acknowledge within the minutes you choose.
- 05 Optionally link the monitor to a component on your status page, so customers see the outage too.
When a check fails in one region, UptimePilot immediately rechecks from the second region. Only when both agree is the monitor marked down and the alert sent. When the site recovers, you get a recovery message with the total downtime.
Multi-region confirmation and false alarms
Most false alarms in website monitoring come from the path between the checker and the site, not from the site itself: a congested route, a DNS resolver hiccup, a brief packet loss at one provider. A single-location monitor turns every one of these into a page at 3 a.m. After a few of those, people mute the channel, and the real outage goes unnoticed.
Requiring a second region to confirm removes that noise. A failure seen from both US East and EU Central is almost always a real problem with the site, its DNS or its certificate. Our article on how to reduce false alarms in uptime monitoring explains thresholds and confirmation in more depth.
What a single check measures
Each HTTP check resolves DNS, opens a TCP connection, performs the TLS handshake, sends the request and waits for the first byte, then reads the body. UptimePilot records the time of each phase. That breakdown is useful when something is slow: a long DNS phase points at the resolver or the DNS host, a long TLS phase at the certificate chain or the server's crypto settings, and a long time to first byte at the application or the database.
Response times are stored per check and summarized as percentiles. The median (p50) shows the typical experience, p95 shows what one visitor in twenty gets, and p99 shows the slow tail that often hides real problems. See website performance monitoring for how to set thresholds on these numbers.
Choosing intervals, timeouts and keywords
A one minute interval means an outage is detected within about two minutes, including confirmation. A 30 second interval cuts that roughly in half, which matters for checkout, login and paid traffic landing pages. Timeouts of 10 to 30 seconds suit most sites. Shorter than 5 seconds tends to flag ordinary slow moments as outages.
Pick keywords that only appear when the page is truly healthy. A product name rendered from the database is better than text in a static header, because a cached shell page can return 200 while the real content is missing.
Uptime math for reference
99.9% availability allows about 43.8 minutes of downtime in a 30.4 day month. 99.95% allows about 21.9 minutes. 99.99% allows about 4.4 minutes. With 30 second checks and two region confirmation, UptimePilot measures downtime with enough resolution to report against all of these targets. Read more in what 99.9 uptime means.
Which plan fits your website monitoring
| Plan | Monitors | Interval | Data history | Monthly, or billed yearly |
|---|---|---|---|---|
| Starter | 50 | 1 minute | 90 days | $49, or $24 per month yearly |
| Team | 200 | 30 seconds | 1 year | $99, or $49 per month yearly |
| Business | 1,000 | 30 seconds | 2 years | $249, or $124 per month yearly |
| Enterprise | 3,000 | 30 seconds | 3 years | $599, or $299 per month yearly |
A small store or a solo developer with a handful of sites fits Starter. A SaaS team that wants 30 second checks and a monthly SLA PDF fits Team. Agencies with many client sites and white-label reports choose Business. Compare every limit on pricing.
01
How often does UptimePilot check my website?
Every minute on Starter and every 30 seconds on Team, Business and Enterprise. Each failed check is rechecked from a second region before an alert is sent.
02
Which regions check my site?
Checks run from US East and EU Central. A monitor is marked down only when both regions confirm the failure, which removes most false alarms caused by network problems along one route.
03
Can I check that specific content is on the page?
Yes. Keyword checks verify that a word or phrase is present, or that it is absent. This catches pages that return 200 but show an error message or an empty layout.
04
Do I need to install anything on my server?
No. Website monitoring works from the outside, the same way a visitor reaches your site. You only enter the URL. For scheduled jobs, heartbeat monitors use a simple request from your job to a unique URL.
05
Are SSL and domain expiry checks included?
Yes, on every plan. UptimePilot warns you before the certificate or the domain registration expires, and flags broken certificate chains.
06
Can I pause alerts during a deploy?
Yes. Maintenance windows stop alerts for the monitors you select during planned work, and checks resume alerting automatically when the window ends.
Know your site is down before your customers do
Add your first URL, choose a keyword and pick who gets the alert. UptimePilot starts checking from two regions right away.