Cloudways review: what $200 a month actually buys

I pay Cloudways roughly $200 a month. That is not a test account set up for a review, and it is not a plan the company gave me. It is the bill for running real production sites for real people who notice when things break. That is the only kind of review worth reading, so it is the only kind worth writing.

The short version

Cloudways sits between you and a raw cloud server. You pick an underlying provider, they run the server, and you get a panel for the things you actually do every week: deploying code, staging changes, taking backups, scaling up when traffic arrives. The parts of server administration nobody misses, patching and firewall rules and certificate renewals, happen somewhere you do not have to look.

After living in that panel for a long time across multiple servers, my verdict is boring and positive. It works, the abstractions hold, and the bill is fair for what it removes from my week. It has real drawbacks, and I will name them, but none of them have pushed me back toward running my own boxes.

Why my bill is $200 a month

The number confuses people, so here is the shape of it. Cloudways bills per server, and I run more than one, because putting every production site on one box means one bad deploy or one noisy neighbour of a plugin can affect all of them at once. Splitting servers costs more and buys blast-radius isolation, and after you have watched one site take down another you never question the spend again.

On top of the servers there are add-ons. The ones worth knowing about: advanced caching and CDN options, and a Cloudflare Enterprise tier that starts at $4.99 per domain per month, which is the kind of pricing that used to require an enterprise contract. Add-ons are optional and they are also how a modest bill becomes a $200 one. Mine is a choice, and it is worth saying plainly that a single small site can sit on the entry tier for a fraction of what I pay.

What Cloudways gets right

The deploy flow is the reason I stay. Git push, the panel builds it, the site updates. When something goes wrong, the previous release is one click away. I have run the alternative, the FTP years, and I will not go back.

Staging is the second reason. Every site gets a staging copy, changes get tested there, and pushing staging live is one click. Cloning a site to a new server, which used to be a weekend, is a form you fill in while making coffee. If you maintain sites for other people, this one feature changes how you work.

Scaling is the third. A server can be resized vertically, more RAM and CPU, without moving anything or re-provisioning. There is a ceiling and horizontal scaling is a bigger conversation, but the common case, a site getting busier, is handled by a slider and a few minutes of downtime-free waiting.

Under all of it: automatic patching, free TLS, an application-level firewall, and monitoring that pages me when something is actually wrong rather than when a metric sneezes. Boring infrastructure is the goal, and mostly it is achieved.

Where it annoys me

No root access by default. There is a way to request it, but the product is built around the idea that you should not need it, and when you genuinely do need something underneath the abstraction, you feel the design decision. For me this comes up rarely. For you it might come up weekly, and you should know that going in.

No email hosting. Cloudways decided, correctly in my view, that managing mail servers is a misery best left to specialists, so you will need Google Workspace or similar alongside. It keeps their focus sharp and adds a line to your costs elsewhere.

The pricing page answers fewer questions than it should. Backup frequency and retention are not stated on the pages I captured in October 2026, and neither is the bandwidth overage rate. You can find out, and support will tell you, but these are exactly the numbers a buyer wants before committing, and their absence from the public pages is a fair criticism.

And the panel, for all its qualities, is a layer. Something like 95 percent of what I want to do has a button. The remaining 5 percent sends me around the abstraction rather than through it, usually involving unusual cron setups or a PHP extension the panel does not expose. It is a tax, it is small, and it is real.

Performance, honestly

I am not going to quote you requests-per-second figures from a benchmark I ran on a Tuesday. Load testing managed WordPress platforms properly, same content, same traffic pattern, repeated runs, is its own piece of work and it lives in our benchmarks section when it lands. What I can tell you from running real sites: page caching plus the CDN add-ons put everyday WordPress performance in the range where visitors stop being the bottleneck and your plugin choices become the bottleneck. That is the honest ceiling of my experience, and anyone quoting more precision from memory is selling something.

Backups, staging, and getting your stuff out

Automated backups run offsite, on-demand backups are there when you want one before a risky change, and restores are a click. Offsite backup storage bills at $0.033 per GB per server, a number so small it has never appeared on my radar as a cost. What I push for, and what you should ask any host for, is a restore you have actually performed. A backup nobody has restored is a hope, and the first time you need one is a bad time to learn the interface.

Moving in is low friction: the first site is migrated free by their engineers, and the migration tooling handles the rest if you bring several. Moving out is possible, the underlying servers are standard, and the data is yours. A host that is easy to leave has to keep earning you, which is the healthiest possible arrangement.

Support

The included support is chat and tickets, and in my usage pattern, which is a developer reading documentation first and opening tickets rarely, it has been fine for what I have needed. There is a paid advanced support tier for people who want faster lanes and more hand-holding. I am deliberately not quoting response times or recounting ticket sagas; I do not keep those numbers, and invented precision is worse than none. Read this as: ordinary support exists, works, and has never been the reason I was unhappy. Also not the reason I am happy. It is infrastructure support.

Pricing and value

You pay per server per month, sized by RAM and CPU, plus add-ons, plus a small offsite backup charge. Entry tiers exist and they are cheap relative to what managed competitors charge for comparable specs, with the usual advice attached: check the current pricing page, because hosting prices move and my October 2026 captures will age.

The value question is simpler than people make it. A raw VPS big enough for the same traffic costs less per month and costs hours per month. If your hours are worth anything, the managed layer pays for itself somewhere around the second time something breaks at 11pm. If you enjoy the operations work, genuinely, then the calculation flips and you should absolutely run your own boxes. Some of the best infrastructure people I know do exactly that. I did it for years. I just stopped wanting to.

Who should use Cloudways

Developers and small teams running several sites who want their hosting boring. Agencies maintaining client work where staging and cloning save whole days. Anyone whose traffic is spiky enough that scaling needs to be easy and calm enough that they never want to SSH in again.

Who should not

People who want root, people who want email included, people optimising for the lowest possible monthly number. If you are those people, look at the comparisons on this site: a raw VPS from DigitalOcean, Vultr or Linode with the management done by you costs less and teaches you more, and I would rather you made that choice knowingly than bought the abstraction and resented it.

Before you commit

Run one non-critical site first. Use the free migration for it, live in the panel for a month, and see whether the abstraction fits the way you work. That month tells you more than any review, including this one.

When you want the head-to-head numbers against SiteGround, Kinsta or a raw VPS, the comparison pages below are where we keep them.