A website isn't finished the day it goes live

Many companies treat their website like a brochure: they pay for it, publish it, and only look at it again when someone says something isn't working. The problem is that a website relies on pieces that change on their own: the server, certificates, dependencies, plugins, browsers, and third-party services.

That's why the worst-case scenario usually isn't a visible outage. It's the form that stopped sending inquiries three weeks ago without anyone noticing. It's the backup that existed on paper but couldn't be restored. It's the domain that expired because the renewal notice went to the email of someone who no longer works at the company.

The good news: most of these problems can be prevented with a short routine. This guide gives you a concrete checklist to review every month, every quarter, and every year, whether your site is custom-built, on WordPress, or on a platform like Shopify or Tiendanube.

Why this is more urgent than a few years ago

Three facts help put it in perspective:

  • WordPress vulnerabilities are growing fast. Patchstack's State of WordPress Security in 2026 report recorded 11,334 new vulnerabilities in 2025, 42% more than in 2024. 91% were in plugins and 9% in themes. WordPress core had only 6, all low priority.
  • The window to react is very short. The same report notes that 46% of vulnerabilities had no patch when they were disclosed, and that for the most targeted ones, the median time to mass exploitation is 5 hours.
  • SSL/TLS certificates keep getting shorter. Following a CA/Browser Forum decision, the maximum validity dropped from 398 to 200 days for certificates issued from March 15, 2026. It will drop to 100 days in March 2027 and to 47 days in March 2029. Renewing them by hand is no longer a reasonable option.

None of this means your site is in danger today. It means that leaving it alone is no longer a neutral way of looking after it.

Monthly checklist: 30 to 60 minutes

1. What brings in business

Start here, because it's what costs the most when it fails and what gets checked the least.

  • Send a real test through every form and confirm the message arrives and doesn't land in spam.
  • Test the WhatsApp buttons, phone numbers, and email links from a phone.
  • If you sell online, place a complete test order, including payment, the confirmation email, and the shipping calculation.
  • Compare this month's conversions with last month's. A sharp drop with no commercial reason is usually a technical problem, not a market one.

2. Updates and security

  • On WordPress: update core, plugins, and themes, in a staging environment first if the site is critical. Remove what you don't use: a deactivated plugin is still code on your server.
  • On a custom build: review dependencies with tools like npm audit or composer audit, and prioritize vulnerabilities that affect code that actually runs.
  • On SaaS platforms: the platform updates itself, but installed apps are still your responsibility. Uninstall the ones you no longer use and review their permissions.
  • Review the list of admin users and remove anyone who should no longer have access, including former employees and previous vendors.
  • Open the Security issues report in Google Search Console. If Google has detected something, that's the first place you'll see it.

3. Backups

A backup that has never been tested is an assumption. Every month, check that copies are being created and that they include both the files and the database.

A practical rule is 3-2-1: three copies, on two different media, with one stored outside the provider that hosts the site. If your hosting has an incident, or you get locked out, a backup that lives only on that same server won't help you.

4. Speed and experience

  • Check the Core Web Vitals report in Search Console, which shows real user data. The “good” thresholds are LCP up to 2.5 seconds, INP up to 200 milliseconds, and CLS up to 0.1.
  • Check what was added during the month. An unoptimized image, a new video, or a campaign pixel can ruin a site that was performing well.
  • Browse the site on a real phone, on mobile data rather than the office Wi-Fi.

5. Search visibility

  • Review the Page indexing report in Search Console to catch new 404 errors or important pages that dropped out.
  • Look at how your main pages are performing. What matters is the trend, not one week's movement.
  • After any major change, confirm that no noindex from the staging environment was left behind, and that robots.txt isn't blocking more than it should.

6. Content

It's the easiest step to skip and the one that speaks most directly to customers. Check prices, opening hours, phone numbers, addresses, team members, expired promotions, and old dates. A “What's new in 2023” on the home page tells every visitor that nobody is watching the site.

Quarterly checklist: a broader look

  • Actually restore a backup. Spin it up in a staging environment and time how long it takes. That's your real recovery time, and it's better to know it before you need it.
  • Audit third-party scripts. Chat widgets, heatmaps, pixels from finished campaigns: each one adds weight, and each one is someone else's code running on your site.
  • Look for broken links and redirects that went wrong after URL changes.
  • Test on several browsers and screen sizes, including Safari on iPhone, which is usually where the surprises show up.
  • Review basic accessibility: contrast, alt text on new images, and keyboard navigation.
  • Check email deliverability. If your forms or your store send emails, verify that SPF, DKIM, and DMARC are still configured correctly.

Yearly checklist: what nobody remembers until it fails

  • Domain ownership. Confirm it's registered to the company, not to a former employee or a vendor. If it's a .com.ar, it's managed through NIC Argentina. Make sure renewal notices go to a company inbox.
  • Centralized access. Build an inventory of hosting, DNS, domain, analytics, Search Console, and payment dashboards, with who holds each login. Ideally, use a password manager and turn on two-factor authentication.
  • Automated certificates. With shorter lifetimes, renewal has to be automatic, and there has to be an alert if it fails.
  • Server versions. Check that your language runtime and database are still supported. For example, according to php.net, PHP 8.2 receives security patches only until December 31, 2026. Versions 8.1 and earlier are already out of support.
  • Costs and vendors. Review what you're paying for hosting, licenses, and apps, and whether you still use all of it. Many subscriptions renew automatically for years without anyone needing them.
  • Legal pages. Update your privacy policy and terms if you've added forms, cookies, payment methods, or new tools.

Who should do what

Maintenance fails less from a lack of knowledge than from a lack of ownership. When everyone thinks someone else is doing it, nobody does. A reasonable split looks like this:

  • Someone on the business side: content, form tests, test orders, and conversion tracking. It doesn't require knowing how to code, especially if the site has a self-managed admin panel.
  • A technical profile, in-house or external: updates, dependencies, backups, server, certificates, and performance analysis.
  • Management or administration: domain ownership, access control, contracts, and renewals.

What matters is putting it in writing, with names and frequency. A recurring task in a shared calendar is enough to get started.

How the routine changes depending on your site

WordPress or WooCommerce: requires the most demanding routine, because the attack surface depends on every installed plugin. Fewer plugins, better chosen and maintained, is the single most effective measure. If you sell online, add a staging environment before every update.

Shopify, Tiendanube, and similar platforms: you offload the infrastructure and platform patches, but you're still in charge of apps, custom themes, content, SEO, speed, and access.

Custom development: with fewer third-party pieces, there are usually fewer urgent updates, but dependencies and the server don't look after themselves. You should have automated monitoring and a clear owner, and the code should be documented so you don't depend on a single person.

Signs that maintenance is no longer enough

Sometimes the checklist reveals a deeper problem. Pay attention if any of these sound familiar:

  • Every update breaks something and nobody dares to touch the site.
  • Speed problems come back as soon as they're fixed.
  • You depend on a single person for any change, even fixing a typo.
  • The site runs on unsupported versions that can't be updated without rebuilding entire parts.
  • You can't access the domain, hosting, or code of your own website.

In those cases, it's worth comparing, with real numbers, how much it costs to keep patching versus how much it costs to rebuild what's broken. You don't always need a new site, but you shouldn't put off that calculation indefinitely either.

Where to start this week

If you've never done any of this, don't try to tackle everything at once. These four actions already remove a large part of the risk:

  • Test your forms and confirm that inquiries are arriving.
  • Check whose name your domain is registered under and when it expires.
  • Ask to see the latest backup, and proof that it can be restored.
  • Schedule a monthly review with a named owner.

If you'd rather have a diagnosis before setting up the routine, at Raven3 we offer a free website audit where we analyze your site's speed, SEO, mobile experience, and conversion, and show you what to tackle first. And if you don't want to carry the technical maintenance yourself, we also offer plans to handle it for you: tell us what you have and we'll take a look.