Rank Math does most of its work in the background — sitemaps, redirects, a 404 log and a Search Console sync. On a normal host all four are quietly competing with your visitors for the same PHP workers.
The visible half is a meta box. The half that needs a host is the crawling, logging and rewriting behind it.
A sitemap index across fifty thousand URLs is a long-running job. Generated inside a page request it hits the execution limit and Google gets a truncated file, or a 500.
The 404 monitor writes a row for every miss. Point a vulnerability scanner at the site for an afternoon and that is tens of thousands of rows nobody will ever read.
A redirect that boots WordPress to answer costs the same as a page view. Moved at the edge, a legacy URL is answered in single-digit milliseconds.
Structured data is assembled from post meta on every load. Cached properly it is free; uncached it is a handful of extra queries on your most-crawled pages.
Googlebot, Bingbot and a dozen SEO tools are real requests. Capacity planning that only counts humans is planning for a fraction of the load.
The failure mode is silent. The sitemap 500s, Search Console keeps showing the last good fetch for a while, and nobody finds out until impressions drop.
It is requested by a crawler rather than a customer, which is exactly why it goes unnoticed until it has been failing for a fortnight.
These differ from a standard WordPress site on the same plan, and each follows from the work the plugin does outside a page view.
| Setting | What we do | Why |
|---|---|---|
| Sitemap generation | Scheduled and cached between builds, not assembled per request | A sitemap index spanning tens of thousands of URLs exceeds the execution limit when built inside a request, and the crawler receives a truncated file rather than an error anyone sees. |
| 404 monitor | Retention window applied and enforced nightly | The monitor writes one row per missed URL, so a single scanner sweep adds tens of thousands of rows that are never read but are backed up and queried forever. |
| Redirect rules | Served at the edge rather than through the plugin | Resolving a redirect through WordPress costs a full bootstrap to emit one header, which is the most expensive way possible to answer a URL that no longer exists. |
| Crawler request budget | Counted as real traffic in worker sizing | Search and SEO-tool crawlers can exceed human sessions on a large catalogue, so capacity planned against analytics alone is planned against the smaller half. |
| Schema and meta queries | Object-cached per post rather than rebuilt each load | Structured data is assembled from post meta on every uncached view, which adds queries precisely to the templates crawlers request most often. |
Every plan includes free migration, daily backups, SSL and 24/7 support.
Free migration, sitemaps that finish, and redirects answered before WordPress wakes up.