Managed Hosting vs. VPS: Who's on Call When Your Website Fails?

Your website crashed at 2 AM. With a VPS, it's your problem; with managed hosting, it's a team's. See the real division of responsibility and cost for each.

by Cleverson Gouvêa

Managed Hosting vs. VPS: Who's on Call When Your Website Fails?

The question of managed hosting vs. VPS rarely arises on the day of purchase. It typically emerges at 2 AM, with your website down, when someone discovers the provider "only supplies the machine". This comparison shows who is responsible for each layer in both models, what each truly costs, and when each choice makes sense for your business.

TL;DR: Managed Hosting vs. VPS in Four Key Points

  • With an unmanaged VPS, the provider is responsible for the hardware and network. The operating system, web server, database, SSL, backups, and security are your responsibility.
  • A 99.99% VPS SLA doesn't cover application failures or configuration errors. Any credit, if applicable, is proportionate to the VPS's value itself.
  • The monthly VPS fee is the smallest part of the cost. What truly adds up are the hours spent on administration and the financial loss from each hour of downtime.
  • A VPS makes sense when you have a technical person with dedicated time and on-call availability. Without this, managed hosting works out cheaper over the year.

What Truly Changes with Managed Hosting vs. VPS

Both models can run on the same type of machine. Often, they literally run in the same data centre. The difference isn't in the hardware: it's in who is responsible for what happens on top of it.

A VPS (Virtual Private Server) is a slice of a physical server, with CPU, memory, and disk reserved for you. You receive an IP, root access, and typically a control panel to switch the machine on, off, and reinstall it. From that point upwards, everything is your responsibility. DigitalOcean states this clearly in their Droplet policies: the machines are "completely unmanaged", and even their support team does not have access to them.

AWS calls this arrangement the shared responsibility model: they look after the security "of" the cloud (building, hardware, virtualisation), and the customer looks after security "in" the cloud, which includes updating and patching the instance's operating system.

With managed hosting, a team assumes these upper layers by contract. They install, update, monitor, perform backups, test restorations, and respond when things break. You remain the owner of your site and data, but you are no longer the one on call.

That's why the managed hosting vs. VPS comparison is rarely about performance. A well-configured VPS is fast. The right question is different: who gets alerted when it goes down?

Who Manages What: The Responsibility Matrix

At Agathas Web, we've operated Linux servers since 2008: Moodle for educational institutions, websites, systems built with Laravel and Next.js, and integrations with the official WhatsApp API. The table below summarises the division of managed hosting vs. VPS, layer by layer. It's the same conversation we have with those who come asking for "just a VPS".

Layer Unmanaged VPS Managed Hosting
Hardware, power, and data centre network Provider Provider (contracted by the managing team)
Operating system and security patches You Managing Team
Web server, PHP/Node, database You Managing Team
SSL certificate and renewal You Managing Team
Firewall, WAF, fail2ban, hardening You Managing Team
Backup and restoration testing You Managing Team
24/7 Monitoring and alerts You Managing Team
Diagnosis when the site goes down You Managing Team
Application code You You (or separate contract)
Content and business data You You

Notice the last two lines. Not even the best managed hosting can fix a bug in your code or a poorly written plugin without a specific agreement for it. Be wary of anyone who promises to "take care of everything" without separating infrastructure from application.

Tasks That Disappear Off the Radar with a VPS

Some responsibilities seem minor until the day they fail:

  • SSL certificates that expire automatically. Let's Encrypt ended expiration notification emails in June 2025. And validity periods will shorten: the default profile drops to 64 days on 10 February 2027 and to 45 days on 16 February 2028, according to the official announcement. Automatic renewal without monitoring is a gamble.
  • Outdated PHP version. PHP 8.2 receives security fixes only until 31 December 2026, according to the supported versions table. Those who don't follow this schedule discover the problem when a plugin stops working or a public vulnerability emerges.
  • Full disk. Logs that grow without rotation, local backups that are never deleted, accumulated sessions. This is one of the most common causes of a frozen database we've encountered.
  • Backup that has never been restored. Having the file proves nothing. Only a tested restore proves it.

Your Website Isn't Covered by the VPS SLA

This is the point that most surprises those comparing managed hosting vs. VPS based on price lists. The VPS provider advertises 99.99% availability, and the customer understands that their website is guaranteed. It isn't.

The DigitalOcean Droplets SLA is a good example because it's clear. It defines unavailability as a loss of connectivity or disk access caused by a failure of DigitalOcean's infrastructure. It expressly excludes problems arising from application code or customer configuration errors.

Translating this into everyday terms:

  1. Did PHP-FPM crash due to insufficient memory? Not covered.
  2. Did MySQL corrupt a table after an abrupt shutdown? Not covered.
  3. Did the certificate expire and the browser block access? Not covered.
  4. Did a brute-force attack bring down the server? Not covered.

And when it is covered, the credit is applied only to the value of the affected machine, on a future invoice. If your VPS costs the equivalent of £40 per month, that's the maximum reimbursement, even if the downtime cost you an entire day's sales.

It's worth calculating how much downtime each SLA level permits in a 30-day month, as it changes how you compare managed hosting vs. VPS in any proposal:

Monthly SLA Permitted downtime per month
99% approx. 7h 12min
99.9% approx. 43min
99.99% approx. 4min

The SLA number matters less than the question: what exactly is covered, and who is on the other end when it breaks?

When the Website Goes Down at 2 AM: Two Scenarios

Imagine the same outage in both sides of the managed hosting vs. VPS equation and see what happens. The database stops responding in the early hours of Saturday morning.

Scenario with Unmanaged VPS

No one is notified because no external alert was configured. The first sign is a customer complaining on WhatsApp at 8 AM. Whoever handles IT (sometimes the owner themselves, sometimes a freelancer) tries to access via SSH, discovers the disk is 100% full, hastily deletes logs, and restarts the service. If the table is corrupted, the search for a backup begins. If the backup was on the same disk, the situation becomes much more challenging.

If the company uses the official WhatsApp API with a webhook hosted on this VPS, messages that arrived during the night may not have been processed. Customer service starts the day without knowing who messaged them.

Scenario with Managed Hosting

External monitoring triggers within minutes. The on-call team receives the alert, identifies the cause, frees up space, restores the service, and logs the incident. In the morning, the client receives a report: what happened, how long it lasted, and what was done to prevent recurrence (log rotation, 80% disk alert, off-server backup).

The difference between the two scenarios isn't technical talent. It's process and on-call availability. A competent person, working alone, sleeps, travels, and gets ill.

Real Cost: The VPS Monthly Fee is the Smallest Part

The classic argument for managed hosting vs. VPS is price. It remains valid but became less clear-cut in 2026. Hetzner, a benchmark for affordable VPS in Europe, adjusted cloud prices from 15 June 2026 for new orders: in Germany and Finland, the CX23 went from €3.99 to €5.49 per month, and the CPX22 from €7.99 to €19.49, excluding taxes and IPv4.

Even so, the machine itself is inexpensive. What isn't cheap is everything else.

What Goes into the VPS Cost

  • Administration hours. Monthly updates, log review, SSL renewal and verification, firewall adjustments, backup testing. Even a simple server requires a few hours per month from someone who knows what they're doing.
  • Separate tools. External storage for backups, uptime monitoring, potentially a WAF.
  • On-call duty. If the website generates revenue, someone needs to be available outside business hours. This comes at a cost, even if it's you.
  • Cost of downtime. Do a simple calculation: monthly online revenue divided by the hours in the month. This is the approximate value of each hour offline, not including reputation damage.

What Goes into the Managed Hosting Cost

A higher monthly fee, with operations embedded. For our managed hosting, the reference range is £250 to £500 per month for institutional websites and blogs, and £600 to £2,500 per month for e-commerce and SaaS, with setup from £800 to £3,500 depending on complexity. Larger operations receive a bespoke quote.

The fair comparison of managed hosting vs. VPS, therefore, isn't "£40 versus £400". It's "£40 + technical hours + tools + risk" versus "£400 with all of that included in the contract".

When a VPS is the Right Choice (and When It Isn't)

There's no universal winner when it comes to managed hosting vs. VPS in all scenarios. We have recommended VPS to clients before and will continue to do so when it makes sense.

Choose a VPS when

  • You have someone on the team with real Linux experience, with dedicated time allocated for it.
  • The project is for testing, staging, or internal use, and a few hours of downtime won't cause financial loss.
  • You need total control over versions and configurations that no managed provider would agree to maintain.
  • The budget is very tight, and the risk of downtime has been consciously accepted.

If this is your situation, start with the security basics. Our Linux server hardening checklist and the guide on how to protect your web server with WAF, fail2ban, and rate limiting cover the minimum an internet-exposed VPS needs on day one.

Don't Choose a VPS when

  • The website or system sells, generates leads, or serves customers, and no one in the company knows what to do if MySQL stops.
  • Your "IT team" is a freelancer who responds when they can.
  • Sensitive personal data is involved. Under UK GDPR, the data controller remains responsible, and the Information Commissioner's Office (ICO) mandates reporting relevant security incidents without undue delay, and within 72 hours of becoming aware where feasible. Discovering a breach a week later is a legal problem, not just a technical one.
  • The platform is critical and has predictable peaks, such as Moodle during exam week or an e-commerce site during a promotional period.

For educational platforms, the comparison gains other variables. We detail this in the article on Moodle on AWS vs. Managed Hosting.

Pitfalls When Hiring Managed Hosting

"Managed" has become a marketing buzzword. There are shared hosting resellers using this name and VPS providers offering a control panel and nothing more. Those comparing managed hosting vs. VPS need to know what they are actually buying. Before signing, ask these questions in writing:

  1. What happens when the website goes down outside business hours? Who receives the alert, and how quickly do they respond? Ask for the figure in the contract.
  2. Is the backup stored off-server? How often is restoration tested? A backup on the same disk doesn't protect against disk failure or ransomware.
  3. Who applies security patches and how quickly after a public vulnerability is disclosed? "When we get around to it" is not an answer.
  4. Does the SLA have consequences? Promising 99.9% without penalties or credit is an intention, not a commitment.
  5. Can I leave with everything? Request access to your data, backups, and environment documentation. Long-term contracts and undocumented environments are the same problem with different names.
  6. Who provides support? A Level 1 agent reading from a script, or someone who has experience operating the type of system you use?

If the answers are vague, you're buying a VPS under a different name and at a higher price.

How Agathas Web Addresses This

Our managed hosting exists for those who want the performance and control of a dedicated server for their project, without becoming its on-call technician. In practice, the contract includes:

  • Environment configured for your stack: WordPress, Laravel, Next.js, Node, Python, Moodle, or bespoke systems, on AWS, Hetzner, Cloudflare, or bare metal, as your project requires.
  • Security by default: Cloudflare WAF, fail2ban, system hardening, automatic SSL, and MFA.
  • Daily and incremental backups, off-server, encrypted, with restoration tested every month.
  • 24/7 monitoring with alerts to the technical team and critical incident response within 15 minutes.
  • Contractual SLA: 99.9% monthly uptime (maximum approx. 43 minutes per month), critical incident resolution within 2 hours, and critical patch application within 72 hours of public disclosure of the vulnerability, with penalties if we fail to comply.
  • Direct support from those who operate the environment, via dedicated WhatsApp, and a monthly report with uptime, backups, and events.

Engagement begins with a free diagnostic of your current infrastructure, delivered within 3 working days. If it makes sense to proceed, we migrate your site, database, emails, and DNS within 7 working days, providing a staging environment for your validation beforehand, and preserving the old environment for 30 days for rollback. The contract is monthly, with no long-term commitment, requiring 30 days' notice to cancel.

And if the diagnostic shows that, in your case, the balance between managed hosting vs. VPS leans towards a simple VPS, we will tell you.

Conclusion: Choose Who Gets Alerted When Your Site Goes Down

Ultimately, the choice between managed hosting vs. VPS is about responsibility. A VPS delivers a good, inexpensive machine and leaves everything else to you. Managed hosting charges more to assume the rest by contract.

If you have a technical person with dedicated time and on-call availability, a VPS might offer the best value for money. If not, the hidden costs will emerge during the first night your site is down.

Want to know which side of the managed hosting vs. VPS debate your business falls on? Request a free diagnostic on the Agathas Web managed hosting page. Within 3 working days, you'll receive an assessment of the risks, costs, and a recommendation, with no obligation.