Site Monitoring
External checks, monitors, and site-assurance signals
External Site Checks
MetricPoints can begin monitoring a website from the outside before you install a browser beacon or WordPress plugin. These checks are read-only and are designed to answer whether the site is reachable, protected, findable, delivering, and behaving as expected from a public visitor-like perspective.
On a site's Health tab, Checks and timing shows when each external check last completed and when it is next due, using relative times. Next due is an estimate: queue time or an earlier recheck can change when the result arrives. Current proof means a check has reported recently; its latest result still needs to be read to tell whether it passed.
Availability and routing
Uptime, expected HTTP status, 200 error/maintenance/parked/challenge documents, TLS/SSL expiry, canonical redirects, redirect-chain limits, important page reachability, and visitor-like browser render health.
Domains and DNS
Domain registration expiry, registrar and nameserver change review, DNS baseline approval, DNS drift, nameservers, A/AAAA/CNAME/MX/TXT/CAA/DS records, DNSSEC, wildcard DNS, resolver disagreement, and sensitive DNS changes.
Search and discovery
Robots.txt, noindex directives, sitemap health, sampled sitemap URLs that are not OK, stale sitemap evidence, canonical targets, AI crawler policy, and llms.txt.
Protection posture
Security-header drift, CSP and CSP-Report-Only presence, risky script evidence, mixed content, public asset checks, and a bounded probe of secret or administrative paths.
Delivery and conversion
Configured form delivery checks, synthetic flows, journey control presence, on-site search smoke, booking-embed slots, analytics collect requests, email-domain health, and important asset integrity. Search smoke checks the homepage and /search for a usable field. Booking embeds are checked only when a booking URL or provider is discovered. A browser worker failure is shown as unavailable, without treating it as a site failure.
Public reputation
Public reputation sources can produce review work. High-confidence malware or safety-provider findings may become critical; single DNSBL-style hits normally need review before being treated as urgent.
How external checks appear in MetricPoints
- Critical now is reserved for likely business-impacting failures such as unreachable sites, broken important journeys, confirmed severe public reputation findings, or credible customer-impacting failures.
- Needs review is used when MetricPoints found evidence an agency should inspect, such as DNS baseline approval, indexability concerns, header drift, single-source reputation hits, or setup decisions.
- Coverage gaps mean MetricPoints does not yet have enough fresh evidence to prove a monitored area is healthy.
- Resolved and archived history remains available as proof without continuing to count as a current site problem.
Check cadence and recovery proof
Routine checks run on the cadence for the site’s monitoring package. When a critical fetch or reachability-style check fails, MetricPoints schedules a faster follow-up so short outages, firewall blips, or restore events can show recovery much sooner than the normal long-running cadence.
Important page discovery
When a site is added, MetricPoints does a short scan of the homepage and common lead or sales paths before creating the asset integrity monitor. It keeps the homepage and reachable important paths it can prove, instead of assuming that every site has a public /contact page. The same scan can be run again from a site’s asset integrity settings.
Alert profiles
A site’s Alerts tab can save the current incident route as a reusable alert profile. Profiles include destinations, severity, recovery notices, and the two critical follow-up milestones. They can then be applied to one site, the current client group, all client-facing sites, or all active sites so agencies do not have to repeat the same notification setup across every client website.
Critical alert follow-up
A confirmed critical incident sends its opening alert immediately. The route can then send one acknowledgement reminder when nobody has taken responsibility and one still-unresolved reminder when the incident remains active. Either milestone can be turned off. MetricPoints does not repeat an alert indefinitely: the maximum is the opening alert plus the enabled milestones, and repeated monitor failures do not create additional alert rounds.
No-install first, deeper evidence later
External monitoring works for WordPress and non-WordPress sites. Adding the browser beacon or WordPress plugin increases confidence and adds evidence, but it is not required to start monitoring public site health.