A homepage is only the beginning
Imagine you run a small agency website. The homepage matters, but so do the contact form entry page, portfolio and service pages that turn a visit into an enquiry. Choosing what to monitor can feel like a second project before you have even created the first check.
PageLantern’s automatic setup helps with that starting point. Give it a public website URL and, when AI recommendations are enabled for your workspace, it proposes a monitoring set for you to review.
Suggestions grounded in your public website
The setup flow inspects public pages and uses observed website links, sitemap entries and source-backed discovery to identify useful targets. Suggestions explain their supporting evidence. A directly inspected page and a page found through a public source are distinguished, so you can see what was actually checked.
The draft can contain up to eight page targets, limited by your remaining monitor slots. Suggestions respect your plan’s available checks and intervals. Extra pages each add a monitor; website health signals for the main target share one monitor.
You decide what becomes a monitor
Open Add website, enter your URL and choose Automatic. Review the suggested pages, remove anything that does not need monitoring, and check the locations and warning thresholds. Test & Start previews checks for the main target before you approve the monitoring set.
That review is useful even when the suggestions look sensible. A contact page might be essential to your business while a rarely used archive is not. Automatic setup supplies a starting point; your priorities determine the final selection.
A suggestion is not proof that a journey works
A public page appearing in a sitemap or search result does not establish its current availability. The preview does not test every optional page, and a recommendation for a checkout page does not prove that a purchase succeeds.
Discovery does not enter private areas or inspect authenticated APIs. A website that blocks the probe may still need an allowlist change before monitoring succeeds. Those limits stay visible instead of being turned into a reassuring but unsupported health claim.
Start small and adjust as you learn
Choose a few pages whose failure would prevent someone from contacting you, using your service or finding key information. Once checks are running, review real failures and tune the monitoring set around what you learn.
Manual setup remains available and makes no AI recommendation request. Workspace owners and admins can manage the AI preference. Regular monitoring continues independently of the recommendation service.