ChangeTrace
Help

Glossary

The words ChangeTrace uses, and what each one actually means.

Anomaly — a day where revenue or orders fell far enough below your 7-day average, on both the size and the unusualness test, to be worth reporting. Detection rules →

Baseline (two meanings) — the trailing 7-day average a day is compared against; and the snapshot of your plugins, theme and versions taken when you connect, against which later changes are detected.

Cause attributable — whether a failure mode is one that a site change could plausibly have caused. Only a WordPress crash and a stuck update qualify. Cause gating →

Confidence — how strongly the timing and correlated signals support a hypothesis. Not a probability that it is correct.

Correlation — the scoring pass that ranks changes from the 24 hours before an incident. Runs every 30 minutes.

Degraded — the site responds, but there is an open incident or a cluster of errors.

Down — proven unreachable: two consecutive failed checks from outside your hosting. The only health state that means your site is actually offline.

Draft — an incident that has just been detected and has no ranked causes yet.

EDD — Easy Digital Downloads, supported as a full equal to WooCommerce.

Event — one recorded thing: a plugin update, a PHP fatal, an order, a heartbeat.

Evidence — measured facts attached to an incident, as opposed to interpretation. Shown with green checks.

Failure mode — how an uptime check failed: WordPress critical error, DNS failure, TLS error, timeout, and so on. The full list →

Heartbeat — the plugin's hourly "still here" message. Its absence means the plugin is not reporting; it says nothing about whether your site is up.

Heartbeat Missing — no heartbeat in two hours. Not an outage.

Hypothesis — a ranked possible cause, with a confidence score.

Incident — a detected problem: a sales drop, or an outage. The two kinds →

Organization — the account that owns your sites, plan and data. Created on first sign-in.

Owner — the organization role that receives incident alert emails.

Performance impact — a before/after page-speed comparison around a change, measured in a lab rather than on real visitors. More →

Safe mode — the plugin's paused state after detecting it is running on a different domain than the one it connected as. More →

Severity — critical (70%+ drop, or site down) or warning (40–69% drop). Only critical emails by default.

Site token — the credential the plugin uses, starting site_tok_. Shown once, stored as a hash, bound to one domain.

Snapshot — the one-time inventory sent when you connect.

Sigma / standard deviation — the measure of how much your daily numbers normally swing. A drop must be at least three of them to count, which is what keeps a naturally spiky shop quiet.

Stale — the internal name for Heartbeat Missing.

UTC day — midnight to midnight UTC. Detection works in these, which is worth remembering if your shop is far from UTC.