# Monitor settings

Per-site settings: what gets crawled, how often, and who hears about it.

## Crawl scope

**Start URL** — where discovery begins, usually the homepage.

**Include and exclude patterns** — path patterns deciding what is in. Excluding is the main lever on both noise and cost, and the usual candidates are:

- Paginated archives (`/page/*`)
- Tag and author archives
- Search result pages
- Anything behind a login
- Staging or preview paths that answer publicly

**Sitemap** — point at yours if it is not at the default location. Without one, pages nothing links to are invisible to the crawler.

**Page limit** — a ceiling, so a runaway URL space (faceted filters generating infinite combinations) cannot consume a crawl.

## Frequency

How quickly you want to know. Daily suits most sites; a documentation site under active edit benefits from it, a brochure site changing quarterly does not need it.

Frequency multiplies pages, and pages are what Monitor is billed on. Crawling everything hourly is usually paying for precision nobody acts on.

## Rendering

Whether to execute JavaScript before checking. Necessary for a site that renders content client-side, unnecessary and slower for one that does not.

If your pages are server-rendered, leaving this off makes crawls faster and cheaper. If you are unsure, turn it on once and compare a page’s rendered output against its source — the difference is the answer.

## Politeness

Request rate and concurrency. The defaults are conservative deliberately: a crawler that hammers a small host looks like an attack and gets itself blocked.

Raise it only if your host is comfortable, and lower it if you see the crawl being rate-limited in [crawl activity](/docs/pages-and-crawl-activity).

## Alerts

Who is notified, through which channel, and at what severity. Set per site — see [alerts and incidents](/docs/alerts-and-incidents).

## Pausing

Pause a site during a migration or a redesign. History is kept, so when you resume the comparison picks up from where it was rather than treating everything as new.

Pausing during a planned change is much better than muting the resulting alerts and forgetting.

## Related

- [Adding a site to Monitor](/docs/adding-a-site-to-monitor)
- [Alerts and incidents](/docs/alerts-and-incidents)
- [Plans and pricing](/docs/plans-and-pricing)