Methodology

How WebsiteCharts calculates uptime, outage frequency, and reliability scores.

Data source

All data comes from WebsiteDown™, which runs HTTP probes against 350+ services every 5 minutes from global infrastructure, and collects community outage reports in real time. WebsiteCharts displays aggregated statistics from that data.

Report windows

Outage reports are aggregated into 5-minute windows. A window with 30 or more reports is classified as an outage event. This threshold filters out isolated user errors and captures genuine service degradation.

Uptime percentage

Uptime is calculated as: (windows with working confirmations) / (total windows with any reports) × 100. A service with no reports in a window is assumed operational. This methodology mirrors industry-standard SLA calculations.

Outage count

Each distinct cluster of consecutive high-report windows is counted as one outage event. A gap of 30+ minutes between spikes creates a new outage event.

Leaderboard ranking

The default leaderboard ranks by total report volume in the selected period. You can re-sort by outage event count or worst uptime percentage using the sort control.

Update frequency

Leaderboard data refreshes every 5 minutes. Individual service history pages cache for 5 minutes with stale-while-revalidate to ensure fast loads.

Questions or corrections?