“Managed hosting” is one of those phrases that means everything and therefore nothing. Every host on the internet uses it, and it rarely comes with a definition. So here is ours, in plain words: managed means four specific jobs, done on a schedule, by a platform that is accountable for them end to end. This post walks through all four.
Backups: copies you can actually restore
A backup is not a thing, it is a habit. Every night, a complete copy of your site — files and database both — is made and moved off the node that serves it, because a backup that lives next to the original shares its fate. Copies are kept for a rolling window, so a mistake discovered on Thursday can be undone from Monday's snapshot, not just from last night's.
The part almost nobody does: restores are rehearsed. A backup you have never restored is a rumour. Periodically we take a real snapshot and bring it back to life in a scratch environment, timing it, checking the database comes up and the images load. When you ask any host about backups, do not ask whether they make them. Ask when they last restored one.
A backup you have never restored is a rumour.
Updates: boring on purpose
Software rots. The CMS core, its plugins, the language runtime underneath, the operating system under that — all of it sprouts security holes and deprecations at its own pace. Managed means someone actually reads the changelogs and applies updates on a schedule, instead of letting a hundred small delays compound into one terrifying upgrade.
The craft is in the order of operations. Security patches go on quickly, out of band. Everything else waits for a calm moment, goes on deliberately, and gets watched afterwards — because the update that breaks your contact form at 4 p.m. on a Friday is only a minor event if somebody is looking at the site at 4:05. Blind auto-update-everything is not management; it is gambling with extra steps.
Monitoring: someone notices
Every minute, an independent check asks your site from the outside: are you up, and how fast did you answer? Underneath that, the nodes themselves are watched — disk filling, memory pressure, certificates approaching expiry. The point of all this plumbing is a single sentence: when something goes wrong, we find out before you do, and long before your customers do.
No honest host promises a site never goes down. The honest questions are: who notices, how fast, and what happens next? Our answers: the platform, within a minute or two, and the fix begins immediately — with the record of it public on our status page, uptime and stumbles alike.
Accountability on the other end
The fourth job is the least technical and the most rare: correspondence that resolves things. When you write in, the answer comes with full context on your site and the authority to act — usually the same day, always in plain language. There is no tier one reading from a script and no ticket number to shepherd through a queue.
This matters more than it sounds. Most hosting problems are small, and small problems are solved fastest with full context and the keys in hand. “Can you make the header photo the new one I attached?” is a two-minute task for the platform that holds root, and a three-day odyssey for a support funnel.
What it costs, and what to ask
Our plans run from $19 to $79 a month, and the difference between them is depth of care, not artificial limits — more watching, more reporting, more hands-on time. Whatever host you end up choosing, us or anyone, take these five questions with you: When did you last restore a backup? Who applies updates, and on what schedule? How do you find out my site is down? How fast do requests actually get resolved? And what happens when I want to leave?
A good host has crisp answers to all five. A great one enjoys being asked. Anyone who bristles has just answered a sixth question you didn't have to say out loud.
Written from the engine room. Questions about anything above? Write to us — answers usually land the same day.