How-to: Set up a monitor
A monitor re-audits a landing page on a schedule (weekly or monthly) and emails you when the score moves beyond a threshold you set. This walks through configuring one.
Monitors require a paid plan. Per-tier limits: Solo Founder = 20; Team = 50; Business = unlimited. Free has no monitors at all. Every paid tier allows weekly and monthly cadences.
1. Open the page workspace
Monitors attach to landing pages, not to URLs in the abstract — so you need a Landing Page record first. From the sidebar, click Landing Pages, then click the row for the URL you want to monitor.
If the URL isn’t in your list yet, add it via + Add page on the same surface, then come back.
2. Go to the Monitor tab
The page workspace has tabs across the top — Dashboard, Audit, Monitor, History. Click Monitor. If a monitor isn’t configured yet, you’ll see an empty-state card with Set up a monitor.
3. Pick a cadence
Two options:
- Weekly — re-audits on a specific day of the week
- Monthly — re-audits on a specific day of the month (1–28)
Pick the cadence that matches how often the page is realistically changing. Weekly makes sense for a page you’re actively iterating on; monthly is plenty for a stable about page.
4. Set the time
For all cadences, pick the hour of day (0–23, in your timezone). Off-hours (early morning) is a sensible default — the email is waiting for you when you start the day, and you don’t pay LLM rate-limit pressure during business hours.
For weekly, also pick day of week (0 = Sunday … 6 = Saturday). For monthly, pick day of month (1–28).
5. Set the alert threshold
The delta threshold is a 1–10 integer that compares the overall rating before and after the run. Two reasonable defaults:
- 2 — alert on any meaningful score swing (recommended for active pages)
- 5 — alert only on dramatic regressions (recommended for stable pages)
Dimension-level swings show up in the email body either way; the threshold gates whether you get the email at all. If you’d rather get a “no change, all good” email every cadence to confirm the monitor’s still running, toggle Notify on no change on.
6. Set the notification email (optional)
By default the alert goes to the email on file for your account. Override with Notify email if you want alerts to go somewhere else — common when an editor configures the monitor but the marketing lead is the one who should see regressions.
7. Save
Click Save monitor. The first run fires according to the schedule (not immediately) — if you want to verify it’s working, you can also click Run now to trigger a one-off run that exercises the same audit + alert pipeline without waiting.
What auto-pause means
A monitor pauses itself in two cases — both visible on the Monitor tab:
- Consecutive failures (3 failed runs in a row, usually scrape errors). Fix the URL or click Resume when ready.
- Plan downgrade — if you downgrade past this monitor’s eligibility, it auto-pauses
with
pause_reason: "plan_downgrade". Re-upgrading auto-resumes.
A paused monitor doesn’t tick and doesn’t burn audit-units.
What’s next
| Goal | Read |
|---|---|
| Understand cadences, thresholds, and pause logic | Concept: Monitors |
| Compare a flagged regression run to its baseline | How-to: Compare audit runs |
| Plan your monitor budget against your audit-unit quota | Plans and pricing |