Skip to content

WordPress maintenance isn’t optional anymore — here’s what we actually do each month

“Website maintenance” is one of those line items that’s hard to argue with and equally
hard to feel good about. You’re paying every month for something that, in a good month,
produces no visible change whatsoever.

That’s a fair thing to be sceptical about. So this is an attempt to answer it properly:
here’s what the work actually consists of, why it can’t reasonably be skipped, and what the
months look like when it’s working.

Why WordPress specifically needs this

WordPress runs a very large share of the web, which is both why it’s a sensible choice and
why it needs attending to. Popularity makes it a target. A vulnerability in a widely used
plugin is worth finding, because the same exploit works on an enormous number of sites.

And a WordPress site isn’t one piece of software. It’s WordPress core, a theme, usually a
page builder, and somewhere between five and thirty plugins — each written by a different
party, each updating on its own schedule, each capable of breaking the others. That’s not a
design flaw; it’s the trade-off that makes the platform flexible. But it does mean the thing
you launched is not the thing you have six months later.

Left alone, a WordPress site doesn’t stay still. It drifts out of date while appearing
completely fine, right up until it isn’t.

What a month actually looks like

Updates, applied in the right order

Core, theme and plugin updates arrive constantly. Some are security patches, some are
feature releases, some are both. The job isn’t clicking “update all” — anyone can do that,
and on a busy site it’s how you find out your checkout page is broken from a customer.

The actual work is: read what changed, decide what’s urgent, apply it on a staging copy
first, check the pages that matter still function, then push it live. Security patches get
handled quickly. Everything else can wait for a scheduled window.

Order matters too. Some plugin updates assume a core version you haven’t installed yet.
Doing them backwards is a reliable way to produce a white screen.

Backups, and testing them

Backups are the boring half of this and the half that saves you. Automated daily backups,
retained far enough back that you can get behind a problem you didn’t notice immediately —
because plenty of problems aren’t noticed immediately.

The part people skip is verification. A backup you’ve never restored is an assumption.
We restore to staging periodically specifically to confirm the restore path works, because
discovering it doesn’t during an actual emergency is a bad day for everyone.

Security monitoring

File integrity checks, failed login monitoring, malware scanning, and watching for the
specific signatures that indicate someone got in — unexpected admin users, modified core
files, new scheduled tasks, injected scripts in the theme.

Most of this catches nothing most of the time. That’s the correct outcome. The value is
in the one month it does, and in catching it in hours rather than whenever Google starts
warning your visitors.

Uptime and performance

Monitoring that actually alerts a person. Sites go down for unglamorous reasons: an expired
certificate, a plugin conflict, a traffic spike, a database that filled up. The difference
between twenty minutes of downtime and two days is almost entirely whether anyone was
watching.

Performance drifts too, usually downward and usually gradually. A new plugin adds its own
stylesheets. Someone uploads a 6MB photo straight off a phone. The database accumulates
revisions and transients. None of it is dramatic; all of it is cumulative. Periodic
housekeeping — image optimisation, database cleanup, caching checks — keeps the site from
quietly getting slower every quarter.

The things that expire

SSL certificates. Domain registrations. API keys. Payment gateway credentials. Third-party
integrations that silently deprecate the version you’re using.

Every one of these fails on a date, often without warning you’d notice. An expired
certificate puts a browser security warning in front of your domain name, which costs you
customers for however long it stands. A lapsed domain is worse. Tracking renewal dates is
tedious, unskilled work, and it is exactly the kind of thing that gets forgotten by a busy
owner and not forgotten by someone whose job it is.

Forms, which break more than anything else

We check them. Monthly, deliberately, by submitting one.

Contact forms fail quietly and constantly. A plugin update changes a setting. A mail
provider tightens its rules and your notifications start landing in spam. A reCAPTCHA key
rotates. The form still looks perfect and still says “thank you” — the message just goes
nowhere.

We have picked up sites that had been dropping every enquiry for months. The owner thought
business was slow. Nothing in this entire list has a worse cost-to-effort ratio than not
checking the form.

The small requests

The practical half of what clients actually use maintenance for: new hours, a price change,
a staff update, swapping a photo, adding a page. Standard updates inside 24 hours.

This part matters more than it sounds, because it’s what keeps a site true. If changing
your own site is a hassle, you won’t, and a site that’s eighteen months out of date is doing
active damage — a customer who finds the wrong hours doesn’t conclude your website is stale.
They conclude you’re closed.

What it looks like when it isn’t happening

Sites we’re asked to rescue tend to arrive in one of a few recognisable states.

The frozen site. Everything is two or three years behind. Nothing is
visibly wrong. Updating is now genuinely risky, because the version jumps are large and the
old theme may not survive them, so the sensible fix is more expensive than three years of
maintenance would have been.

The compromised site. Someone got in through a known vulnerability in an
outdated plugin. The patch had been available for months. Cleanup means removing injected
code, checking for persistence, rotating credentials, and often requesting review from Google
if the site got flagged — all of it costing considerably more than the patch would have.

The slow bleed. Nothing broke. The site just got heavier, slower and
further out of date. Rankings drifted. Conversions drifted. No single incident to point at,
which is exactly why nobody addressed it.

None of these involve anyone being careless. They involve a business owner having a
business to run.

Should you just do it yourself?

Genuinely, you can. If you’re comfortable with a staging environment, you read plugin
changelogs, you have tested backups and you’ll notice within a day if something breaks —
do it yourself. We’d rather you did that than pay for something you don’t need.

The honest question isn’t capability. It’s whether it will still be happening in month
fourteen, when you’re busy, when nothing has gone wrong in a year and the whole exercise
feels unnecessary. That’s the month it matters, and that’s the month it stops.

Maintenance is one of those costs that looks pointless right up until the moment it looks
cheap.

How we handle it

Our maintenance and support runs
on the sites we host, which means staging environments, tested backups and the update path
are all in one place rather than coordinated across three vendors. Standard update requests
go out inside 24 hours.

If you’re not sure what state your site is currently in, that’s a reasonable place to
start — we’ll look at what’s running, what’s behind, and whether your form actually works.
Usually at least one of those answers is a surprise.

Ask us to check your site →

Newsletter

Signup our newsletter to get updated information, and insight about the technology

Latest article

Get Started

Name(Required)
Interested in:(Required)