Hosting Fundamentals

What Is Server Uptime?

A plain-English guide to how uptime works, what the percentages really mean, why it matters for your website, and which UK hosts deliver the best reliability.

Server Uptime in One Sentence

Server uptime is the percentage of time your website's server is switched on, working, and accessible to visitors — think of it like the opening hours of a shop, except your “shop” should be open 24/7/365.

When a hosting company promises “99.9% uptime,” they're saying your site should be reachable for all but roughly 43 minutes each month. That sounds impressive — until you realise 43 minutes of downtime on a busy Monday morning could mean hundreds of lost visitors, missed sales, and damaged trust. This guide explains exactly what those numbers mean, what causes downtime, how to monitor it, and which UK hosts actually deliver on their promises.

How Does Server Uptime Work?

Step 1

Your website lives on a server

When someone types your domain name, their browser connects to a physical (or virtual) server that stores your files and delivers them. If that server is running and reachable, your site is "up." If not, visitors see an error page.

Step 2

The host keeps the server running 24/7

Your hosting provider is responsible for power, cooling, network connectivity, hardware maintenance, software updates, and security patches — everything needed to keep that server online around the clock.

Step 3

Uptime is measured as a percentage

Over a given period (usually a month), uptime is the proportion of time the server was accessible. 99.9% uptime over a month means roughly 43 minutes of downtime. 100% uptime is essentially impossible — even the biggest tech companies experience outages.

Step 4

Hosts promise a guarantee (SLA)

Most hosts advertise an uptime SLA (Service Level Agreement) — typically 99.9% or 99.95%. If they fall below that, you can usually claim account credit. The SLA is a promise, not a measurement — actual uptime may differ.

What Do the Numbers Actually Mean?

Each extra “nine” makes a dramatic difference. Here's what the most common uptime levels translate to in real-world downtime:

Uptime %Downtime / MonthDowntime / YearGrade
99%~7 hours 18 min~3 days 15 hoursPoor
99.5%~3 hours 39 min~1 day 20 hoursBelow Average
99.9%~43 minutes~8 hours 46 minIndustry Standard
99.95%~21 minutes~4 hours 23 minGood
99.99%~4.3 minutes~52 minutesExcellent
99.999%~26 seconds~5 minutesEnterprise-Grade

⚠️ Key insight: The jump from 99.9% to 99.99% cuts annual downtime from ~8 hours 46 minutes to just ~52 minutes. That extra “nine” typically requires significantly better infrastructure — cloud hosting, redundant hardware, and automatic failover.

What Causes Downtime?

Hardware Failure

High Impact

Disk crashes, memory faults, power supply failures, or overheating. Redundant hardware (RAID, dual PSUs) mitigates this, but nothing is failure-proof.

Software & OS Updates

Medium Impact

Server restarts during kernel patches, PHP upgrades, or control panel updates. Usually planned and brief, but sometimes extended by complications.

Network Issues

High Impact

Fibre cuts, BGP routing errors, DNS outages, or upstream provider failures. These can affect entire data centres regardless of server health.

DDoS Attacks

High Impact

Malicious traffic floods overwhelm the server or its network. Hosts with DDoS mitigation recover faster; those without can go offline for hours.

Traffic Spikes

Medium Impact

Unexpected surges in legitimate traffic can exhaust CPU, RAM, or bandwidth on shared or undersized servers — essentially a self-inflicted overload.

Human Error

Medium Impact

Misconfigured firewalls, accidental deletions, or botched deployments by either the host's team or the site owner. One of the most common causes.

Data Centre Outages

Critical Impact

Power grid failures, cooling system breakdowns, or natural disasters (floods, fires). Tier III+ data centres have redundancy to minimise this risk.

Shared Resource Exhaustion

Medium Impact

On shared hosting, another site on the same server consuming excessive resources can slow or crash your site — the "noisy neighbour" problem.

Why Uptime Matters (and What to Watch Out For)

Why High Uptime Matters

Revenue Protection

Every minute of downtime costs money — particularly for e-commerce sites. A 99.99% host gives you roughly 52 minutes of downtime per year vs 8+ hours at 99.9%.

SEO & Rankings

Google factors site availability into rankings. Frequent or prolonged downtime can cause your pages to be de-indexed and your rankings to drop significantly.

User Trust

Visitors who encounter errors lose confidence. A study by Hosting Tribunal found 88% of users are less likely to return to a site after a bad experience — downtime counts.

Email Reliability

If your hosting also manages email, server downtime means you stop receiving emails. Messages can bounce, and you may never know what you missed.

Data Integrity

Sudden crashes (especially during write operations) can corrupt databases. Hosts with high uptime typically have better hardware and backup systems.

Peace of Mind

Not constantly worrying about whether your site is online lets you focus on growing your business rather than firefighting hosting issues.

The Realities to Keep in Mind

100% Is Impossible

No host can guarantee absolute zero downtime. Even AWS, Google Cloud, and Azure experience outages. Marketing claims of "100% uptime" should be taken with scepticism.

SLAs Have Loopholes

Most uptime guarantees exclude "scheduled maintenance," DDoS attacks, and third-party failures. Read the fine print — the guarantee may not cover what you expect.

Compensation Is Minimal

Typical SLA credits are 5-10% of your monthly bill per hour of downtime. If you pay £5/month and your site is down for 2 hours, you might get 50p back — it won't cover lost sales.

Higher Uptime Costs More

The jump from 99.9% to 99.99% requires significantly more redundancy, failover systems, and infrastructure — which is reflected in the price. Budget hosts rarely achieve it.

Measuring Is Tricky

Your host's internal monitoring may show 99.99% while third-party tools record 99.8%. Measurement location, frequency, and methodology all affect the number.

Planned Downtime Still Counts

Even if your host warns you about maintenance windows, your site is still inaccessible to visitors during those periods. "Planned" doesn't mean "painless."

How to Monitor Your Website's Uptime

Never rely solely on your host's own monitoring. Use an independent tool to verify their claims and catch outages before your visitors do.

ToolPriceCheck IntervalBest For
UptimeRobotFree (50 monitors)Every 5 minBest free option for small sites
PingdomFrom $10/moEvery 1 minBest for detailed performance analysis
Better UptimeFree (10 monitors)Every 3 minBest for teams needing status pages
Hetrix ToolsFree (15 monitors)Every 1 minBest for monitoring from multiple locations
StatusCakeFree (10 monitors)Every 5 minBest UK-based monitoring service

💡 Our recommendation: Start with UptimeRobot (free, 50 monitors, 5-minute checks). It's more than enough for most sites and takes about 2 minutes to set up. Upgrade to a paid tool only if you need 1-minute checks or real-user monitoring.

Who Needs to Prioritise Uptime?

High uptime is critical if you…

  • E-commerce stores where every minute of downtime means lost sales
  • Business websites that serve as the primary customer touchpoint
  • SaaS applications with paying users who expect 24/7 access
  • Media or news sites with time-sensitive content and high traffic
  • Client projects where your reputation depends on site reliability

Standard uptime is fine if you…

  • Personal hobby blogs where occasional downtime has no real cost
  • Staging or development environments used only by your team
  • Temporary campaign or event pages with a short lifespan
  • Sites with extremely low traffic and no revenue dependency

UK Hosting Provider Uptime Comparison

How the most popular UK-accessible hosts compare on uptime guarantees and real-world performance:

ProviderSLA GuaranteeMeasured Uptime
SiteGround99.9%99.99%+
Kinsta99.9%99.99%+
Cloudways99.99%99.98%+
IONOS99.9%99.95%+
Krystal99.9%99.96%+
Fasthosts99.9%99.9%+
20i99.9%99.95%+
A2 Hosting99.9%99.93%+
BluehostNot published99.9%+
DreamHost100%99.95%+
HostArmada99.9%99.94%+
Scala Hosting99.9%99.92%+

💡 Tip: Measured uptime figures come from third-party monitoring and community reports. They represent typical performance but can vary. Always verify with your own monitoring tool after signing up.

Related Guides

Server Uptime — Frequently Asked Questions

What does 99.9% uptime actually mean?
99.9% uptime means the server is expected to be accessible for 99.9% of a given period. Over a month (30 days), that allows for roughly 43 minutes of downtime. Over a year, it's about 8 hours and 46 minutes. Most budget-to-mid-range UK hosts target this level.
Is 99.9% uptime good enough?
For personal blogs, small business brochure sites, and most WordPress sites, 99.9% is perfectly adequate. For e-commerce stores processing orders 24/7 or critical business applications, you should aim for 99.95% or higher — hosts like Kinsta and Cloudways typically deliver this.
What is the difference between uptime and availability?
They're often used interchangeably, but technically "uptime" means the server is powered on and running, while "availability" means the server is both running AND accessible to users. A server can be "up" but unreachable due to a network issue — so availability is the more meaningful metric.
How can I check my website's uptime?
Use a free monitoring tool like UptimeRobot or StatusCake. They ping your site every few minutes from multiple locations and alert you (via email, SMS, or Slack) when your site goes down. Most offer dashboards showing your historical uptime percentage.
Does my hosting provider monitor uptime for me?
Most hosts monitor their own infrastructure internally, but they rarely share real-time data with customers. Some premium hosts like Kinsta and Cloudways provide uptime dashboards. For independent verification, always use a third-party monitoring tool.
What should I do when my site goes down?
First, check if it's truly down (use a tool like "Down For Everyone Or Just Me"). Clear your DNS cache and try a different browser/device. If it's confirmed down, contact your host's support immediately. Document the start time — you may need it for SLA claims.
Can I get compensation for downtime?
If your host offers an uptime SLA and fails to meet it, you can usually claim account credit. The process typically involves opening a support ticket with evidence (monitoring logs showing downtime). Compensation is usually modest — 5-10% of your monthly fee per hour of downtime.
Does shared hosting have worse uptime than VPS?
Not necessarily. The uptime guarantee is often the same (99.9%). However, shared hosting is more susceptible to the "noisy neighbour" effect — another site on your server consuming excessive resources. VPS hosting isolates your resources, making performance more consistent.
What is a "five nines" uptime guarantee?
"Five nines" means 99.999% uptime — just 5 minutes and 15 seconds of downtime per year. This is enterprise-grade and requires massive redundancy (multiple data centres, automatic failover, redundant everything). It's typically only available from major cloud providers like AWS and Google Cloud.
Is 100% uptime realistic?
No. Even the world's largest tech companies (Google, Amazon, Microsoft) experience outages. DreamHost advertises a 100% uptime guarantee, but this is a marketing commitment — they compensate you for any downtime, not a claim that downtime never occurs. Always plan for some level of outage.

Want Reliable Hosting?

Our comparison tool analyses uptime, speed, support, and value across 23 UK-accessible hosts — helping you find a provider that keeps your site online.

Last updated April 2026 · Based on testing of 23 UK hosting providers · Written for beginners · Affiliate disclosure