Documentation

Scheduled scans

Monitoring for important URLs with intervals, regions and alert rules.

What scheduled scans are

Scheduled scans are the monitoring side of turbometrics. Instead of measuring only occasionally, you check important URLs regularly and notice changes earlier.

That makes sense for:

  • home pages
  • landing pages
  • shops and category pages
  • conversion pages
  • client projects under ongoing support

Why scheduled scans are often more valuable than single measurements

A single scan only answers the question: what did the page look like just now?

Scheduled scans additionally answer:

  • does the page stay stable?
  • does an area degrade after updates?
  • are there recurring problems?
  • when did an anomaly first occur?

That is what makes history, baseline and alerts genuinely useful in the first place.

Scan intervals per plan

The available interval depends on the active plan:

Interval Plan
Weekly Free
Daily Free, Starter
Every 12 hours Pro
Every 6 hours Pro
Hourly Agency

Shorter intervals make sense for pages where performance problems need to surface quickly — shops or heavily frequented landing pages, for example.

Important settings per target

For every target, the central monitoring decisions can be made:

  • interval
  • region
  • alert mode (auto or custom)
  • alert on a failed scan
  • email on critical alerts
  • email on warnings
  • webhook on critical alerts
  • webhook on warnings

These settings apply per target and make the difference between plain monitoring and genuinely usable signals.

Alert mode: auto or custom

For every target you can choose how alerts should be triggered.

Auto mode

turbometrics detects anomalies automatically, based on changes compared to the previous scan:

  • Score drop: if the score falls by 10 or more points → warning; 20 or more points → critical
  • TTFB: more than doubled compared to the last scan → warning; exceeds 3 sec. → critical
  • TBT: exceeds 600 ms → warning; exceeds 1,200 ms → critical
  • HTTP status: if the page does not respond with HTTP 200 → critical
  • Visual regression: screenshot deviates by ≥ 15% from the last healthy capture → warning; ≥ 30% → critical (if enabled, see below)
  • Scan failure: if the scheduled scan fails → critical (if enabled)

Auto mode suits pages for which you have no fixed target values but still want to be informed about sudden regressions.

Custom mode

You define thresholds for each metric yourself — one each for warning and critical:

Metric Warning example Critical example
Score below 75 below 60
TTFB above 800 ms above 1,500 ms
LCP above 2,500 ms above 4,000 ms
TBT above 600 ms above 1,200 ms
Visual deviation above 15% above 30%

Leaving a field empty means: this threshold is disabled. So you can, for example, set only a critical value without enabling a warning.

This mode suits you if you have concrete performance targets and want to control exactly when an alert is triggered.

Visual regression detection

In addition to the performance metrics, turbometrics can detect when the appearance of a page has changed substantially — because of a broken CSS deployment, for example. To do so, every scan compares the screenshot with the last capture that showed no anomaly.

  • Has to be enabled per target via the "Check visual regression" checkbox — disabled by default.
  • Desktop and mobile are checked separately and trigger independent alerts.
  • As long as a visual alert is open, the comparison reference stays frozen, so that a second broken scan is not mistakenly treated as "normal".

→ For details on thresholds and auto/custom mode, see Alerts.

Test alert

Every scheduled scan has a Send test alert button in its settings. It lets you check whether email and webhook notifications are delivered correctly — without waiting for a real scan.

What the test alert triggers:

  • an email, if Email on critical is enabled for this target
  • a webhook, if Webhook on critical is enabled and a webhook URL is stored in the profile

The test alert is delivered as a clearly marked test signal and creates no real alert entry in the alert history. If none of the notification options is active, a corresponding note appears.

Tips on thresholds

  • For a shop with good hosting the values can be stricter; for a small blog more generous
  • Thresholds that are too tight lead to frequent alerts, which are quickly ignored
  • Better to start with softer values and tighten them after a few weeks of history data

A typical start

For new projects, less is often more. A good start usually consists of a few core pages:

  • the home page
  • the most important category or landing page
  • a heavily frequented subpage
  • an area relevant to transactions

Later you can extend the set via sitemap import or bulk actions.

Plan-dependent limits

The maximum number of scheduled scans depends on the active plan. When the limit is reached, turbometrics blocks the creation of new targets cleanly with a notice instead of failing silently.

Further features such as warning notifications or webhooks can be plan-dependent as well.

Plans & limits

Common pitfalls

  • monitoring too many pages too early
  • enabling warning notifications for pages of little relevance
  • enabling webhooks even though no webhook URL is stored in the profile
  • forgetting important regions or core pages
  • setting thresholds too tight, which leads to too many alerts

Further reading

Sitemap importBulk actionsAlertsPlans & limits