Straight answers. Including the limits.

AlwaysUp checks your site from the outside, the way visitors see it. Whatever the cause (a host outage, a traffic flood, a fatal error, a database crash), we confirm the failure from several independent locations so a blip or a slow page never triggers a switch. About a minute in, visitors see a verified saved copy of your pages on your own domain. A few minutes later, a fully working copy of your WordPress site takes over: logins, forms and checkout all work. When your host is back, everything that changed is brought back to your server and checked before traffic returns, and an AI incident report tells you what happened.

It's real. For the first minute or so visitors see your saved pages. Then a working WordPress copy of your site runs on our infrastructure, on your domain: logins, forms, checkout, comments and your custom plugins work. It's a working site, not a picture of one. We promise it works and stays online; we don't promise it's exactly as fast as your own server.

Fast. Your saved pages go up about a minute after your site goes down, once the failure is confirmed from several places. The fully working copy of your site takes over within about four to five minutes of the crash. We're measuring real timings on real hosts during early access and will publish them.

It does the reading and explaining for you. After every incident it writes a plain-English report with the likely root cause. When an update crashes your site it names the plugin or theme responsible and offers a one-click fix. It checks pages before they go into your saved copy, so an error page dressed up as "OK" never gets served. It walks you through setup, writes your readiness report, explains slowdowns and traffic spikes, and answers questions about your sites in plain language. Every AI answer is labelled as AI-generated.

For a moment, possibly. We deliberately confirm a failure before acting, so a hiccup or a slow page doesn't flip your site. That takes about a minute. And if a form or order can't be delivered during a switch, the visitor gets an honest "please try again", never a fake "thank you".

Yes. On the backup site your WooCommerce store keeps working: customers browse, add to cart, check out and pay. Every order taken during the outage is brought back to your server afterwards and verified, and your AI incident report tells you what you sold while your host was down.

They're saved continuously and brought back to your server. Before traffic returns, we archive your server's own state, copy everything over, and verify it table by table and file by file. If that check can't pass safely, the backup site simply keeps serving until it can.

No. Your host stays exactly where it is. Install the plugin, connect, and point one DNS record to us. AI tells you exactly what to click at your DNS provider. No API keys, no storage accounts, no server configuration and no tickets to your host.

You copy one record (some sites two) into your DNS settings and paste it. That's it. The AI setup assistant tells you exactly where to click for your provider, and a live check turns green the moment it works. There's virtually zero downtime while you switch, and you can switch back any time. If your provider can't do what's needed, we can host your DNS for you and you change it once. We check your mail records first and keep them working; if we can't prove your email stays safe, setup stops and tells you exactly why.

Almost every site on any host, but not all, and we say so up front. We can't protect a site yet if:

• nobody can get into the domain's DNS settings (for example, a former developer holds the login),
• DNSSEC is switched on and can't be changed,
• a company's IT team won't allow the DNS change, or
• the host's firewall blocks our traffic (we expect this to be rare).

The free readiness check tells you before you pay anything.

Yes, any time. Your site never moves: it stays on your host, under your domain. Leaving means changing one DNS record back, and you're exactly where you started.

Offsite, away from your host. Each backup is encrypted on your own site before it leaves, with a key unique to your site. After the first full upload, only what changed is sent, so it stays light even on large sites. Backups adapt to your server's load and pause before they could slow your site down. How often depends on your plan, and your dashboard always shows your actual latest recovery point.

Every night we restore a sample of backups on separate machines and record the result. Every restore is checked over the web afterwards and rolled back automatically if anything fails. You can also undo a restore yourself for 72 hours.

No, and we won't pretend otherwise: AlwaysUp isn't a security firewall. What we do is help you recover: restore your site to a clean, verified point, onto your current server or a brand-new one, and undo the restore if you need to.

It works on its own, with no account. When a bad plugin or theme update causes a fatal error, it shows visitors your last good page instead of an error, and AI names the culprit and offers a one-click fix. It also gives you an AI readiness report (what would happen if your host died right now, plus three specific recommendations), keeps a local last-known-good copy of your site and sends basic uptime alerts.

Not yet. These are early-access prices, per site, per month. Plans and limits may change before launch, and we'll tell you before anything changes for you.

We never switch off a working site automatically when an allowance ends. You'll get clear notices and the option to extend.

We're in early access. Join the list and we'll get in touch as we open places. We'd rather bring sites on carefully than quickly.

No. AlwaysUp is an independent product built for WordPress sites. It isn't affiliated with or endorsed by the WordPress Foundation.