What about uptime and backups

Availability is monitored continuously and backed by a published commitment; backups run automatically on a daily and weekly schedule for every plan, including the smallest order. Neither is sold as an upgrade. What follows is what those two sentences actually mean in practice, and what they still leave for you to do.

Why availability matters differently here

On an ordinary brochure site, an outage at three in the morning is mostly invisible. On an estate built for search, it is not. Crawlers work continuously and do not retry politely for long. A site that is unreachable when a crawler arrives has spent that visit on an error page, and repeated across an estate that becomes a pattern of unreliability rather than an unlucky moment.

That is the argument for continuous monitoring rather than periodic sampling. Sampling tells you the platform was up when you checked. Monitoring tells you whether it was up when your visitors and the crawlers arrived, which is the only version of the question that matters.

What is automated for you

• Daily and weekly backups of files and databases, on the platform, without configuration.
• Restores through the control panel, at whatever granularity you need rather than all or nothing.
• Continuous service monitoring, with support staffed around the clock rather than in office hours.

What remains yours

This is the part worth reading twice, because it is where people are caught out:

• Testing the restore. A backup nobody has restored is an assumption. Restore something deliberately, early, while it does not matter.
• An off-platform copy of anything irreplaceable. Platform backups protect against the ordinary failures. They are on the same platform, and a copy elsewhere is what protects against the unusual ones.
• Application-level integrity. A backup faithfully preserves a database that a bad plugin update corrupted. Knowing when the corruption started is your job, not the backup's.

A habit worth adopting

Before any change you would describe as significant, take an out-of-band copy and know how you would put it back. That covers the case the schedule cannot: the change you made at ten in the morning that you want to undo at eleven, before the next scheduled run.

What the availability figure covers and excludes is set out in Uptime and Backups. If you would rather own the whole restore strategy yourself, a VPS or dedicated server puts it in your hands along with the operating system.