Website Maintenance

Website maintenance services for the failures that never show an error

A site can be online, fast and reporting 99.98% uptime while its contact form has delivered nothing for three weeks. Everything that matters most fails quietly, returns a healthy response code, and waits for a customer to tell you.

Monitoring built around the things that earn money, not around whether the server answered the door.

The basics

What are website maintenance services?

Website maintenance services are the ongoing work of keeping a live site secure, current and actually functioning: applying updates through staging rather than in production, monitoring the paths that generate revenue, testing backups by restoring them, tracking expiry dates before they arrive, and fixing the small breakages that accumulate into a site nobody trusts.

Most retainers sold under this name are an update schedule with a monthly PDF attached. Updates matter, but they are the part a plugin can do. The part that needs a person is noticing when something has stopped working without announcing it.

The core problem

Uptime is the least useful thing you monitor

Almost every maintenance arrangement reports uptime, because it is trivial to measure. A service requests your homepage every few minutes and records whether it answered. That produces a reassuring number and tells you close to nothing.

Uptime confirms that a server responded to a request for one page. It does not confirm that the form on that page sends anywhere, that checkout completes, that the payment gateway is authorising, that search returns results, or that the version of the page a logged-out visitor gets is the same one you are looking at while logged in.

Every failure worth worrying about returns a 200. A form that silently stops delivering, a broken checkout step, an expired API key, a page rendering empty because a plugin update removed a field. All of them look identical to an uptime monitor: the server responded, so the site is up.

This is why maintenance conversations so often start with a business saying enquiries have dropped and nobody knowing for how long. The monitoring was pointed at the wrong question.

Failure modes

The failures that never show an error

These are the ones we find most often on sites that have been running without real maintenance. None of them announce themselves.

Forms that stop delivering

The most expensive silent failure there is. A deliverability change, an expired integration or a plugin conflict, and the form still shows a success message while the email goes nowhere. Enquiries do not bounce, they simply never arrive.

Backups nobody has restored

A backup you have never restored is a hypothesis. Common outcomes when it is finally tested: the database is included but uploads are not, the archive is corrupt, or restoring takes eleven hours nobody had budgeted.

Expiry dates nobody owns

SSL certificates, domain renewals, API credentials, paid plugin licences. Each is a dated event that will happen, and each tends to be owned by whoever set it up rather than by anyone still working there.

Revenue paths untested

Checkout, booking, quote requests and account signup are the pages that pay for the site, and they are almost never monitored beyond whether they load. Load is not the same as complete.

Analytics that stopped recording

A tag removed during an unrelated change, or a consent banner update that blocks everything. You lose the data and the ability to notice you lost it, which is how a three-month gap gets discovered at year end.

Abandoned dependencies

Plugins and libraries whose authors stopped publishing updates. Nothing breaks on the day they are abandoned, which is exactly the problem: the risk accumulates invisibly until a disclosure makes it urgent.

Two of these you can check yourself right now. Our bulk PageSpeed checker runs several templates at once, which is how performance drift shows up as a pattern rather than a one-off, and the SEO audit tool will surface structural breakages that have crept in since launch.

Process

How our website maintenance services run

Maintenance works on a rhythm rather than a ticket queue, because the point is to find things before anyone reports them.

Continuous

Synthetic checks that complete a real journey rather than requesting a page: submit the form and confirm the message arrived, add to cart and reach the payment step, log in. Alerts go out when the journey fails, not when the server does.

Weekly

Updates reviewed and applied through staging, error logs read, form submissions reconciled against what the business actually received, and any new console or checkout errors triaged.

Monthly

A restore performed from backup into a scratch environment to prove it works. Performance re-measured across key templates. Expiry calendar reviewed for anything falling due in the next 90 days.

Quarterly

Dependency audit for abandoned or unused components, accessibility spot check, and a plain summary of what broke, what was fixed and what we recommend changing structurally.

The monthly restore is the item most often dropped from maintenance retainers and the one we will not drop. It is the only way the word backup means anything.

Deliverables

What our website maintenance services include

Journey monitoring

The paths that generate revenue, checked end to end on a schedule, with alerting that reaches a person rather than an unread inbox.

Form delivery verification

Test submissions sent and confirmed as received at the destination, so a silent delivery failure surfaces in days rather than in a quarterly revenue review.

Staged updates

Core, plugin, theme and dependency updates applied to a staging copy, checked, then promoted. Production is never the place where an update is tried for the first time.

Tested backups

Offsite, versioned, and restored monthly into a scratch environment with the restore time recorded so you know what an actual recovery costs in hours.

Security monitoring

Vulnerability disclosures tracked against your specific installed inventory, file integrity monitoring, and a documented path for when a component has no patch available.

Expiry calendar

Certificates, domains, licences and API credentials with named owners and renewal dates, reviewed every month so nothing expires on a weekend.

Performance drift tracking

Key templates measured on a schedule against the numbers recorded at launch, because sites do not get slow overnight, they get slow one plugin at a time.

Included fix time

A monthly allowance for small changes and breakages, so minor work happens instead of accumulating into a list that eventually justifies a rebuild.

Plain reporting

What broke, what was fixed, what is at risk and what we recommend. Written to be read by whoever pays for it, not padded with graphs.

Honest answer

When you do not need a maintenance retainer

Retainers are sold far more often than they are needed. Some sites genuinely do not warrant one.

A static brochure site

Flat HTML with no forms, no logins and no database has very little that can silently fail. Certificate renewal and an occasional content edit is the whole job, and that is not a retainer.

A fully managed platform with no apps

A Shopify store on a stock theme with no third-party apps has most of this handled by the platform. The moment you add apps or customise the theme, that stops being true.

You have the capability in-house

If someone technical already owns updates, staging and restore testing, you need our documentation more than our time. We are happy to set up the monitoring and hand it over.

The site is being replaced shortly

If a rebuild is already scheduled, keep the current site secure and patched and put the money into the new build instead of into maintaining something with a known end date.

Investment

How much do website maintenance services cost?

Pricing follows how much can go wrong, which means dependency count and whether the site takes money, not how many pages it has.

Brochure site

$150 to $350 a month. Staged updates, tested backups, uptime and journey monitoring on forms, expiry calendar and an hour of fix time. Suits marketing sites where the form is the conversion.

Content or lead generation site

$350 to $800 a month. Adds performance drift tracking, security monitoring against your installed inventory, monthly restore testing and two to four hours of included changes.

eCommerce

$800 to $2,000 a month. Checkout and payment journeys monitored continuously, integration and API watch, priority response, and a documented rollback path for a failed update during trading.

Worth comparing against the alternative honestly. A single silent form failure lasting a month costs most businesses more than a year of the first tier, which is the entire commercial argument for maintenance and the reason we lead with monitoring rather than with updates.

Common questions about website maintenance

Your host maintains the server, not your site. Managed hosting typically covers the operating system, the web server and often nightly backups of the files as they exist. It does not test whether your contact form delivers, whether checkout completes, or whether a plugin update broke a template. Those are application concerns and they sit outside what hosting is responsible for.

They handle the applying, which is the easy half, and they introduce a risk of their own: an update that breaks a template at two in the morning stays broken until someone looks. Automatic updates are reasonable for security patches on well maintained components. They are not a substitute for a staging step, and they are no help at all with the failures that never produce an error.

Usually you would not, which is what makes it the most costly failure on the list. The form still displays a success message because the site accepted the submission. The break is further along, in delivery. The only reliable detection is a scheduled test submission checked at the destination, plus reconciling the count of submissions the site recorded against what the business actually received.

New features, redesigns and content strategy. A retainer covers keeping what exists working, plus a modest allowance for small changes. If you want new templates, a restructured site or ongoing campaigns, those are projects and should be scoped and priced as projects rather than absorbed into a monthly fee that then quietly stops covering maintenance.

Monthly, by restoring one into a scratch environment and recording how long it took. Testing matters more than frequency of backup, because the common failures only appear on restore: uploads excluded from the archive, a database dump that will not import, or a recovery that takes far longer than the business assumed. An untested backup is a plan, not a safeguard.

That depends on the tier and it should be written down before you need it. Our eCommerce tier includes priority response because a store down on a Saturday is losing money by the hour. For a brochure site, next business day is usually the honest and appropriate answer, and paying for more than that is paying for reassurance rather than for risk.

Yes, and most of what we maintain we did not build. It starts with an audit, because we will not take responsibility for a site we have not looked inside. Occasionally that audit finds something that needs fixing before a retainer makes sense, and we would rather say that at the start than discover it during an incident.

Find out what is already failing quietly

We will submit your forms and check they arrive, complete a purchase or enquiry end to end, look at what is installed and unmaintained, and check whether your backups can actually be restored. Most reviews turn up at least one thing nobody knew about. No obligation, no sales pitch.

Get my free site review
  • Form delivery tested
  • Backups verified
  • Dependencies audited