What a 99.9% uptime guarantee is actually worth

Nine hours of permitted downtime, a refund you have to claim yourself, and definitions of "downtime" that exclude most of what breaks. Reading the SLA properly.

Every host advertises an uptime figure and several advertise a "100% uptime guarantee", which is on its face an impossible promise. Both are marketing artefacts, and the gap between the number and the protection is wide enough to be worth a few minutes' reading.

What the percentages permit

So the ubiquitous 99.9% permits nearly nine hours of outage a year without breaching anything. For most small sites that is genuinely fine. For a shop taking orders on a Saturday, nine hours is a bad day with a number attached.

The three clauses that hollow out the guarantee

1. You have to claim it. SLA credits are almost never applied automatically. You must notice the outage, document it, and open a ticket — often within a short window, sometimes as little as seven days. Hosts are entirely aware that hardly anyone does this.

2. The credit is trivial. The typical remedy is a pro-rata refund of the affected period, or in the better cases a month's hosting fee. If your £6/month site is down for four hours during your busiest trading day, the compensation is a few pence — or at best £6. It does not price the loss; it prices the hosting.

3. The definition of "downtime" is narrow. This is the important one. Most SLAs measure whether the server responds to a network-level check. Excluded, typically: scheduled maintenance, DDoS attacks, problems caused by your own code or plugins, third-party failures, and anything the host classifies as force majeure. A site that responds in 30 seconds instead of 300 milliseconds is not "down" under most definitions, and a database error page returned with an HTTP 200 may not be either.

So what is a "100% uptime guarantee"?

It means the SLA promises credit for any measured downtime rather than only downtime beyond a threshold. It is a billing policy, not a claim about physics. 20i, Namecheap, DreamHost and Hyve all offer versions of this, and it is a mildly positive signal — a company that expects frequent outages would not choose that structure. Just do not read it as a promise of never going down.

Where an SLA does mean something

At the enterprise tier. Hyve's 100% SLA sits alongside a named technical account manager, 24/7 UK engineers, a 60-second answer target, and contractual remedies negotiated per customer rather than published on a pricing page. That is a real commercial commitment, and it is priced accordingly at £150/month and up.

Similarly, DigitalOcean's 99.99% Droplet SLA is a meaningful statement about their infrastructure — but note it covers the server being reachable, and everything you install on it is your responsibility entirely.

Measure it yourself

The most useful response to all of this is to stop relying on the host's own reporting. Set up independent monitoring — UptimeRobot's free tier gives you five-minute checks, Better Stack and Pingdom do more — and configure it to check for a specific string in your page content, not merely an HTTP 200. That catches the failure mode that matters most: a site that is technically up and serving a database error, an expired certificate warning, or a white screen.

Then, if you ever do need to claim a credit, you have your own timestamps rather than an argument about the host's.

The honest framing: uptime guarantees are a signal of intent, not insurance. If downtime genuinely costs you money, the protection is redundancy and independent monitoring, not a clause. If it costs you mild embarrassment, 99.9% from a decent host is entirely adequate and the SLA is irrelevant either way.

Published 30 July 2026. Prices and ownership change — check the comparison table for current figures, and the provider's own site before buying. No affiliate links anywhere on this site.